Abstract
To identify symptom phenotypes among breast cancer survivors using latent profile analysis (LPA) and examine differences in fear of cancer recurrence (FCR) and social support levels across phenotypes. A multicenter cross-sectional study recruited 357 breast cancer survivors through convenience sampling from three university-affiliated tertiary hospitals in Beijing between November 2024 and January 2025. Symptoms were assessed using the Memorial Symptom Assessment Scale (MSAS), while the Fear of Cancer Recurrence Inventory–Short Form and Social Support Rating Scale measured FCR severity and perceived social support. LPA was performed using severity scores of nine MSAS symptoms to determine the optimal symptom profile model. Three symptom phenotypes were identified: Emotion-reactive (33.9%), Somatic-burden fatigue–pain (13.2%), and Regulated low-symptom (52.9%). Significant differences among these phenotypes were observed in medical burden, treatment method, activities of daily living, and illness cognition (P < 0.05). FCR levels were significantly higher in the Emotion-reactive and Somatic-burden phenotypes compared with the Regulated low-symptom phenotype (p < 0.001, η2p = 0.199), with no difference between the two high-symptom phenotypes. After adjusting for covariates, differences in perceived social support across symptom phenotypes were not significant (p = 0.119, η2p = 0.012). Breast cancer survivors exhibit substantial heterogeneity in symptom phenotypes: those with Emotion-reactive and Somatic-burden fatigue–pain phenotypes experience higher FCR levels, whereas individuals with the Regulated low-symptom phenotype show better symptom control and comparatively lower FCR levels. Integrating symptom phenotypes with FCR may aid rapid identification of high-risk subgroups and support precision nursing in survivorship care.
Supplementary Information
The online version contains supplementary material available at https://doi.org/10.1038/s41598-026-58532-9.
Keywords: Breast cancer survivors, Symptom phenotypes, Latent profile analysis, Fear of cancer recurrence, Social support
Subject terms: Cancer, Diseases, Health care, Oncology, Risk factors
Introduction
Breast cancer is the most common invasive cancer among women worldwide and the fourth leading cause of cancer-related death1. In recent years, advances in widespread screening and standardized treatment have substantially improved survival among patients with breast cancer2,3. As survival has increased, clinicians have gradually shifted their focus toward long-term follow-up, with the aim of effectively managing the long-term physical and psychosocial consequences of cancer and its treatment, which has become a key domain in contemporary frameworks for survivorship care quality4. Breast cancer patients are affected not only by the tumor itself but also by multiple symptoms arising from anti-tumor treatments5. Among these, sleep disturbance, pain, fatigue, anxiety and depression, fear of cancer recurrence, and cognitive impairment are particularly common6,7, and their impact may persist for several years after treatment completion. These symptoms markedly reduce quality of life, undermine adherence to rehabilitation-phase treatments, and consequently, hinder improvements in oncological outcomes8,9.
Among breast cancer survivors, symptom burden often manifests as multiple co-occurring and interrelated physical and psychological symptoms that form symptom clusters10,11. Symptom clusters exert negative synergistic effects; 50–70% of breast cancer patients experience a moderate to severe symptom cluster burden after treatment, which severely impairs their ability to perform activities of daily living (ADL)12,13. However, breast cancer patients receiving chemotherapy are usually hospitalized in short cycles and spend most of the inter-cycle periods at home, where they lack access to professional symptom management.
Symptom clusters in breast cancer patients undergoing cancer treatment have been widely investigated, with most studies focusing on the associations among symptoms to determine the number and types of clusters. However, symptom profiles vary substantially among patients with different sociodemographic and clinical characteristics14. In recent years, an increasing number of studies have adopted person-centered approaches, such as latent profile analysis15, which is a mainstream person-centered analytic method used to identify distinct symptom phenotypes or patterns of symptom cluster characteristics in oncology populations. This approach helps to elucidate the characteristics and needs of patient subgroups and to implement precision cancer care, which may in turn improve the allocation of healthcare resources. Clinically, identifying symptom phenotypes may help clinicians recognize survivors at higher risk of persistent symptom burden or psychological distress and provide more targeted follow-up, education, and supportive interventions. However, few studies to date have specifically examined symptom phenotypes among Chinese breast cancer survivors and their associations with fear of cancer recurrence and social support. Integrating these constructs is clinically relevant because symptom phenotypes reflect different patterns of physical and psychological symptom burden, which may shape survivors’ recurrence-related threat appraisal. Persistent or distressing symptoms may be interpreted as possible signs of cancer progression or recurrence, thereby contributing to FCR. Recent evidence in breast cancer has shown significant interactions between fear of cancer recurrence (FCR) and symptoms, with FCR being positively associated with most symptoms16. Social support may function as a buffering factor by helping patients interpret symptom experiences more adaptively, reduce uncertainty, and access emotional and informational resources, thereby potentially weakening the link between symptom burden and FCR17. In addition, recent psycho-oncology research has emphasized that survivorship outcomes are also shaped by psychological constructs such as self-management capacity, body image, and awareness of bodily sensations. These factors may influence how survivors interpret symptoms, cope with uncertainty, and respond to the perceived threat of recurrence. Related intervention studies have further suggested that patient-centered communication, skill transfer for health management, and emerging digital approaches such as virtual reality may help improve emotional well-being, body perception, and fear of cancer recurrence in breast cancer survivors18,19.
In this study, we used latent profile analysis (LPA) to identify subgroups of breast cancer survivors characterized by distinct symptom phenotypes and examined the effects of different symptom phenotypes on FCR and social support. We hypothesized that breast cancer survivors would exhibit heterogeneous symptom phenotypes and that those with more severe emotional or somatic symptom burden would report higher levels of FCR and lower levels of perceived social support. Accordingly, the aims of this study were to identify distinct symptom phenotypes among breast cancer survivors and to compare FCR and social support across these phenotypes. The findings may provide evidence for identifying high-risk survivor subgroups and informing phenotype-based symptom management strategies in breast cancer survivorship care.
Materials and methods
Study design and setting
This multicenter cross-sectional study was conducted in three university-affiliated tertiary general hospitals in Beijing, China. These hospitals admit breast cancer patients from different regions across China, which may increase the diversity of the study sample. Between November 2024 and January 2025, eligible patients with breast cancer were recruited using a convenience sampling method during routine clinical visits in the oncology and breast surgery units across the three hospitals. The researchers explained the study purpose and procedures and assured participants that their information would be kept confidential. Patients who agreed to participate provided written informed consent before independently completing the questionnaire using Wenjuanxing, a widely used Chinese WeChat-based online survey platform.
Participants
The inclusion criteria were as follows: (1) pathologically confirmed diagnosis of breast cancer, (2) age ≥ 18 years, (3) clear consciousness and ability to communicate independently, and (4) awareness of the diagnosis and provision of written informed consent. The exclusion criteria were as follows: (1) presence of a psychiatric disorder or cognitive impairment, (2) hearing or speech impairment, (3) distant metastasis of cancer, and (4) severe physical illness or other malignant tumors requiring treatment.
Sample size
The sample size for this study was determined based on Monte Carlo simulation studies and methodological reviews of latent profile analysis15,20. Previous research has indicated that when identifying 3–4 latent classes, with medium between-class differences (Cohen’s d for mean indicator differences ≈ 0.5) and the smallest class proportion ranging from 10 to 15%, a total sample size of approximately 300 is sufficient to achieve high class recovery accuracy (correct classification rates mostly above 0.80) and stable parameter estimates21. Accordingly, we anticipated identifying approximately three symptom phenotypes among breast cancer survivors and assumed that the smallest phenotype would account for at least 10% of the sample; therefore, the required sample size was set at a minimum of 330 cases. Considering an anticipated 10% attrition or missing questionnaires, we planned to recruit at least 365 participants.
Measures
Sociodemographic information included age, marital status, educational attainment, and perceived medical financial burden. Clinical data comprised cancer stage, treatment regimen, time since cancer diagnosis, and patients’ self-rated understanding of their disease.
Psychological and symptom-related variables were assessed using three standardized instruments. (1) Symptom burden was evaluated using the simplified Chinese Memorial Symptom Assessment Scale (MSAS)22,23, which assesses the frequency, severity, and distress of common cancer-related symptoms over the past week. It includes 28 physical and 4 psychological symptom items, scored from 0–4 or 1–4, with higher scores reflecting greater overall symptom distress. In this study, the Cronbach’s α for the MSAS was 0.753. Symptom prevalence across all 32 MSAS items was calculated. LPA indicator selection is described in the Data analysis Sect. (2) Fear of cancer recurrence was measured with the Fear of Cancer Recurrence Inventory Short Form (FCRI-SF)24. The scale comprises 9 items rated on a 5-point Likert scale (0–4), yielding a total score of 0–36, with higher scores indicating more severe FCR. The Chinese version has demonstrated excellent internal consistency (Cronbach’s α = 0.912)25, and Cronbach’s α in the present study was 0.718. (3) Perceived social support was assessed with the Social Support Rating Scale (SSRS)26, a 10-item scale covering objective support, subjective support, and support utilization. The total score ranges from 0 to 66, with higher scores indicating higher levels of social support; Cronbach’s α in the original studies ranged from 0.89 to 0.94 and was 0.804 in the current sample.
Ethical considerations
This study was approved by the Ethics Committee of Beijing Shijitan Hospital, Capital Medical University (IIT2024-135-003), and all participants were fully informed about the study procedures and provided written informed consent.
Data analysis
Statistical analyses were performed using R version 4.4.1, with the tidyLPA package for latent profile analysis, and the dplyr and ggplot2 packages for data management and visualization. Symptom phenotypes among breast cancer survivors were identified using the severity scores of nine symptoms from the MSAS (i.e., Difficulty concentrating, Difficulty sleeping, Dry mouth, Feeling sad, Lack of appetite, Lack of energy, Numbness/tingling, Pain, and Worrying) as indicators. To balance model parsimony and clinical interpretability and to reduce the risk of overparameterisation given the sample size (n = 357), we prespecified the criteria for LPA indicator selection. First, we calculated the prevalence of all 32 MSAS symptoms (Supplementary Fig. 1) and, based on prior literature27, considered symptoms with a prevalence ≥ 30% as candidates. From these candidates, we prespecified nine MSAS symptom severity items (Difficulty concentrating, Difficulty sleeping, Dry mouth, Feeling sad, Lack of appetite, Lack of energy, Numbness/tingling, Pain, and Worrying) as indicators to capture clinically salient symptom domains while maintaining an interpretable profile solution. Models with different numbers of latent classes were fitted, and candidate models were compared based on the Bayesian information criterion (BIC), the adjusted Bayesian information criterion (ABIC), and entropy. Classification quality was assessed using average posterior probabilities for most likely class membership (AvePP), summarised as the minimum–maximum range across classes for each candidate solution. The optimal symptom phenotype model was selected by jointly considering these indices, the proportion of each latent class, and clinical interpretability. After identifying the symptom phenotypes, the LPA-derived classes were used as grouping variables. Contingency tables and chi-square (χ2) tests were applied to compare differences in sociodemographic and disease-related characteristics across phenotype groups. One-way analysis of variance was used to examine the associations between symptom phenotypes and levels of fear of cancer recurrence and social support. For outcomes with significant overall tests, analysis of covariance (ANCOVA) models were fitted, adjusting for age, tumour stage, and perceived medical financial burden, and eta squared or partial eta squared were reported as effect size indices. For outcomes with significant overall tests, Tukey post hoc comparisons were performed to describe differences among symptom phenotypes.
Results
Participants’ sociodemographic and clinical characteristics
Three hundred fifty-seven breast cancer survivors were included in this study. All participants were women, with a mean age of 54 years. Most were married (87.4%) and had one child (62.2%). Half of the participants had stage II breast cancer (51.5%). Regarding time since cancer diagnosis, 36.7% of participants were diagnosed within 3 months, 15.4% within 3–6 months, 16.2% within 6–12 months, and 31.7% more than 12 months previously. In addition, 35.6% of patients were undergoing chemotherapy and 23.8% were receiving endocrine therapy (Table 3).
Table 3.
Sociodemographic and clinical characteristics across three symptom phenotypes among breast cancer survivors (N = 357).
| Variable | No. (%) | χ2 | P | |||
|---|---|---|---|---|---|---|
| Total (n = 357) | Emotion-reactive symptom phenotype (C1, n = 121) | Somatic-burden fatigue–pain phenotype (C2, n = 47) | Regulated low-symptom phenotype (C3, n = 189) | |||
| Age | 6.658 | 0.354 | ||||
| < 40 | 29 (8.1) | 10 (8.3) | 5 (10.6) | 14 (7.4) | ||
| 40–50 | 114 (31.9) | 45 (37.2) | 16 (34.0) | 53 (28.0) | ||
| 51–60 | 95 (26.6) | 28 (23.1) | 8 (17.0) | 59 (31.2) | ||
| > 60 | 119 (33.3) | 38 (31.4) | 18 (38.3) | 63 (33.3) | ||
| Marital Status | 1.153 | 0.979 | ||||
| Unmarried | 18 (5.0) | 6 (5.0) | 3 (6.4) | 9 (4.8) | ||
| Married | 312 (87.4) | 107 (88.4) | 41 (87.2) | 164 (86.8) | ||
| Divorced | 14 (3.9) | 4 (3.3) | 1 (2.1) | 9(4.8) | ||
| Widowed | 13 (3.6) | 4 (3.3) | 2 (4.3) | 7 (3.7) | ||
| Education Level | 4.720 | 0.787 | ||||
| Junior high school or below | 48 (39.7) | 19 (40.4) | 71 (37.6) | |||
| High school | 138 (38.7) | 24 (19.8) | 12 (25.5) | 51 (27.0) | ||
| Associate degree | 87 (24.4) | 24 (19.8) | 8 (17.0) | 35 (18.5) | ||
| Bachelor’s degree | 67 (18.) | 22 (18.2) | 6 (12.8) | 30 (15.9) | ||
| Master’s degree or above | 58 (16.3) | 3 (2.5) | 2 (4.3) | 2 (1.1) | ||
| Employment Status | 4.060 | 0.852 | ||||
| Employed | 7 (2.0) | 32 (26.4) | 12 (25.5) | 47 (24.9) | ||
| Unemployed | 91 (25.5) | 28 (23.1) | 10 (21.3) | 36 (19.0) | ||
| Retired | 74 (20.7) | 47 (38.8) | 20 (42.6) | 89 (47.1) | ||
| Laid-off | 156 (43.7) | 1 (0.8) | 0 (0.0) | 0 (0.0) | ||
| Other | 1 (0.3) | 13 (10.7) | 5 (10.6) | 17 (9.0) | ||
| Childbearing Status | 3.816 | 0.702 | ||||
| No children | 24 (6.7) | 10 (8.3) | 4 (8.5) | 10 (5.3) | ||
| One child | 222 (62.2) | 78 (64.5) | 28 (59.6) | 116 (61.4) | ||
| Two children | 92 (25.8) | 26 (21.5) | 14 (29.8) | 52 (27.5) | ||
| Two or more | 19 (5.3) | 7 (5.8) | 1 (2.1) | 11 (5.8) | ||
| Living Environment | 1.210 | 0.546 | ||||
| Urban | 281 (78.7) | 92 (76.0) | 36 (76.6) | 153 (81.0) | ||
| Rural | 76 (21.3) | 29 (24.0) | 11 (23.4) | 36 (19.0) | ||
| Medical Payment Method | 8.854 | 0.355 | ||||
| Self-paid | 7 (2.0) | 3 (2.5) | 1 (2.1) | 3 (1.6) | ||
| Medical insurance | 276 (77.3) | 89 (73.6) | 35 (74.5) | 152 (80.4) | ||
| Public medical care | 6 (1.7) | 2 (1.7) | 1 (2.1) | 3 (1.6) | ||
| NRCMS | 67 (18.8) | 27 (22.3) | 9 (19.1) | 31 (16.4) | ||
| Commercial insurance | 1 (0.3) | 0 (0.0) | 1 (2.1) | 0 (0.0) | ||
| Personal Monthly Income | 7.721 | 0.259 | ||||
| < 2000 | 109 (30.5) | 44 (36.4) | 11 (23.4) | 54 (28.6) | ||
| 2001–5000 | 139 (38.9) | 36 (29.8) | 23 (48.9) | 80 (42.3) | ||
| 5001–10,000 | 82 (23.0) | 31 (25.6) | 9 (19.1) | 42 (22.2) | ||
| > 10,000 | 27 (7.6) | 10 (8.3) | 4 (8.5) | 13 (6.9) | ||
| Medical Burden | 20.031 | 0.003 | ||||
| No burden at all | 66 (18.5) | 15 (12.4) | 6 (12.8) | 45 (23.8) | ||
| Minimal burden | 120 (33.6) | 36 (29.8) | 11 (23.4) | 73 (38.6) | ||
| Moderate burden | 132 (37.0) | 52 (43.0) | 25 (53.2) | 55 (29.1) | ||
| Severe burden | 39 (10.9) | 18 (14.9) | 5 (10.6) | 16 (8.5) | ||
| Cancer Stage | 4.582 | 0.598 | ||||
| 0 | 10 (2.8) | 3 (2.5) | 2 (4.3) | 5 (2.6) | ||
| Ⅰ | 100 (28.0) | 37 (30.6) | 14 (29.8) | 49 (25.9) | ||
| Ⅱ | 184 (51.5) | 63 (52.1) | 19 (40.4) | 102 (54.0) | ||
| Ⅲ | 63 (17.6) | 18 (14.9) | 12 (25.5) | 33 (17.5) | ||
| Family History of Cancer | 1.700 | 0.427 | ||||
| No | 272 (76.2) | 97 (80.2) | 34 (72.3) | 141 (74.6) | ||
| Yes | 85 (23.8) | 24 (19.8) | 13 (27.7) | 48 (25.4) | ||
| Surgical Method | 11.791 | 0.299 | ||||
| No surgery | 56 (15.7) | 23 (19.0) | 6 (12.8) | 27 (14.3) | ||
| Breast-conserving surgery | 61 (17.1) | 19 (15.7) | 12 (25.5) | 30 (15.9) | ||
| Modified radical mastectomy | 100 (28.0) | 30 (24.8) | 13 (27.7) | 57 (30.2) | ||
| Radical mastectomy/Extended radical mastectomy | 68 (19.0) | 30 (24.8) | 6 (12.8) | 32 (16.9) | ||
| Total mastectomy | 69 (19.3) | 19 (15.7) | 10 (21.3) | 40 (21.2) | ||
| Sentinel lymph node biopsy | 3 (0.8) | 0 (0.0) | 0 (0.0) | 3 (1.6) | ||
| Cancer Diagnosis Time(months) | 9.900 | 0.129 | ||||
| < 3 | 131 (36.7) | 56 (46.3) | 12 (25.5) | 63 (33.3) | ||
| 3–6 | 55 (15.4) | 14 (11.6) | 7 (14.9) | 34 (18.0) | ||
| 6–12 | 58 (16.2) | 19 (15.7) | 10 (21.3) | 29 (15.3) | ||
| > 12 | 113 (31.7) | 32 (26.4) | 18 (38.3) | 63 (33.3) | ||
| Treatment Method | 23.782 | 0.008 | ||||
| Radiotherapy | 6 (1.7) | 0 (0.0) | 2 (4.3) | 4 (2.1) | ||
| Chemotherapy | 127 (35.6) | 44 (36.4) | 11 (23.4) | 72 (38.1) | ||
| Endocrine therapy | 85 (23.8) | 29 (24.0) | 13 (27.7) | 43 (22.8) | ||
| Targeted therapy | 43 (12.0) | 9 (7.4) | 11 (23.4) | 23 (12.2) | ||
| Follow-up during rehabilitation period | 29 (8.1) | 6 (5.0) | 5 (10.6) | 18 (9.5) | ||
| Postoperative | 67 (18.8) | 33 (27.3) | 5 (10.6) | 29 (15.3) | ||
| Follow-up Frequency | 1.885 | 0.984 | ||||
| Once every 3 months | 90 (25.2) | 27 (22.3) | 12 (25.5) | 51 (27.0) | ||
| Once every 6 months | 38 (10.6) | 12 (9.9) | 5 (10.6) | 21 (11.1) | ||
| Once a year | 16 (4.5) | 5 (4.1) | 2 (4.3) | 9 (4.8) | ||
| No follow-up | 3 (0.8) | 1 (0.8) | 0 (0.0) | 2 (1.1) | ||
| No follow-up during the treatment period | 210 (58.8) | 76 (62.8) | 28 (59.6) | 106 (56.1) | ||
| ADL | 15.214 | 0.004 | ||||
| Fully independent | 273 (76.5) | 83 (68.6) | 31 (66.0) | 159 (84.1) | ||
| Needs assistance from family | 82 (23.0) | 37 (30.6) | 15 (31.9) | 30 (15.9) | ||
| Completely dependent on others | 2 (0.6) | 1 (0.8) | 1 (2.1) | 0 (0.0) | ||
| Level of Understanding of the Disease | 10.755 | 0.029 | ||||
| Fully understand | 140 (39.2) | 34 (28.1) | 18 (38.3) | 88 (46.6) | ||
| Partially understand | 199 (55.7) | 80 (66.1) | 26 (55.3) | 93 (49.2) | ||
| Do not understand at all | 18 (5.0) | 7 (5.8) | 3 (6.4) | 8 (4.2) | ||
NRCMS New Rural Cooperative Medical Scheme.
Symptom phenotypes identified using LPA
Using the severity scores of the nine MSAS symptoms as indicators, latent profile models with 2 to 4 classes were fitted (Table 1). As the number of latent classes increased, both BIC and ABIC showed decreasing trends, with the 4-class model having the lowest values (BIC = 8624.387; ABIC = 8472.108), indicating the best statistical fit. The entropy values were high across models (2-class: 0.942; 3-class: 0.898; 4-class: 0.916), suggesting good classification precision. Consistent with the entropy results, the AvePP range indicated good classification precision across candidate models (2-class: 0.980–0.982; 3-class: 0.857–0.993; 4-class: 0.884–0.974). In terms of class proportions, the 3-class model yielded C1, C2, and C3 proportions of 33.9%, 13.2%, and 52.9%, respectively, whereas in the 4-class model, C1 accounted for only 5.3%. As it is generally recommended that profile sizes should be at least 5–8% of the sample28,29, this class was considered very small. Considering the model fit, class proportions, and clinical interpretability, the 3-class model was deemed the optimal symptom phenotype solution for subsequent analyses.
Table 1.
Fit indices for latent profile models based on nine symptoms among breast cancer survivors (N = 357).
| Model | BIC | ABIC | Entropy | AvePP range | Latent profiles proportion | |||
|---|---|---|---|---|---|---|---|---|
| C1 | C2 | C3 | C4 | |||||
| 2 | 8806.767 | 8717.938 | 0.942 | 0.980–0.982 | 0.359 | 0.641 | – | – |
| 3 | 8728.834 | 8608.280 | 0.898 | 0.857–0.993 | 0.339 | 0.132 | 0.529 | – |
| 4 | 8624.387 | 8472.108 | 0.916 | 0.884–0.974 | 0.053 | 0.126 | 0.294 | 0.527 |
BIC Bayesian Information Criterion, ABIC Sample-size Adjusted Bayesian Information Criterion, AvePP range Range of Average Posterior Probabilities for Most Likely Class Membership.
Table 2 presents the mean and standard error values of the nine symptom severity scores across the three symptom phenotypes, and Fig. 1 depicts the corresponding symptom severity profiles. Based on the symptom patterns, three latent profiles were interpreted and labeled. The Emotion-reactive symptom phenotype (C1 33.9%, n = 121) showed a moderate overall symptom burden, with pronounced peaks in emotion-related symptoms such as Worrying and Feeling sad. The Somatic-burden fatigue–pain phenotype (C2 13.2%, n = 47) was characterized by a broad burden of predominantly somatic symptoms, with the highest scores for Lack of energy, Difficulty sleeping, Pain, and Numbness/tingling, whereas emotional symptoms were relatively mild. The Regulated low-symptom phenotype (C3 52.9%, n = 189) had the lowest scores across all nine symptoms, mostly around or below 0.5 points, suggesting only mild and scattered symptom experiences.
Table 2.
Derived symptom phenotypes and symptom severity among breast cancer survivors (N = 357).
| Symptom | Emotion-reactive symptom phenotype (C1, n = 121) | Somatic-burden fatigue–pain phenotype (C2, n = 47) | Regulated low-symptom phenotype (C3, n = 189) |
|---|---|---|---|
| M ± SE | M ± SE | M ± SE | |
| Difficulty concentrating | 0.81 ± 0.09 | 0.98 ± 0.15 | 0.33 ± 0.05 |
| Difficulty sleeping | 1.33 ± 0.11 | 1.77 ± 0.23 | 0.65 ± 0.07 |
| Dry mouth | 1.07 ± 0.09 | 1.45 ± 0.16 | 0.60 ± 0.06 |
| Feeling sad | 2.25 ± 0.05 | 0.34 ± 0.07 | 0.12 ± 0.02 |
| Lack of appetite | 0.54 ± 0.09 | 0.62 ± 0.15 | 0.41 ± 0.06 |
| Lack of energy | 1.20 ± 0.09 | 1.89 ± 0.16 | 0.51 ± 0.05 |
| Numbness/tingling | 0.63 ± 0.09 | 1.32 ± 0.19 | 0.38 ± 0.05 |
| Pain | 0.98 ± 0.09 | 1.60 ± 0.16 | 0.34 ± 0.04 |
| Worrying | 2.22 ± 0.08 | 1.53 ± 0.15 | 0.46 ± 0.06 |
Fig. 1.
Symptom severity profiles of the three latent symptom phenotypes.
Differences in sociodemographic and clinical characteristics among breast cancer survivors with different symptom phenotypes
Table 3 presents the differences in sociodemographic and clinical characteristics of the participants, categorized by symptom phenotype. The results showed that medical burden, treatment method, ADL, and level of understanding of the disease differed significantly across symptom phenotypes (P < 0.05).
Differences in FCR and social support across symptom phenotypes among breast cancer survivors
Table 4 shows the differences in FCR and social support across the different symptom phenotypes among breast cancer survivors. After controlling for age, tumor stage, and medical burden, the three phenotypes differed significantly in total FCR scores (p < 0.001, η2p = 0.199), indicating a medium effect size. The Regulated low-symptom phenotype had the lowest FCR level (9.95 ± 0.62), which was significantly lower than that of the Emotion-reactive symptom phenotype (C1 15.48 ± 0.68) and the Somatic-burden fatigue–pain phenotype (C2 15.19 ± 0.90), whereas the difference between C1 and C2 was not significant (Tukey: C3 < C1 ≈ C2). With respect to social support, after adjustment for covariates the overall differences among symptom phenotypes were no longer significant (p = 0.119, η2p = 0.012), and the mean scores were similar across the three classes (C1: 39.13 ± 0.99, C2: 39.49 ± 1.32, C3: 40.99 ± 0.90). Tukey posthoc tests did not detect any significant pairwise differences. In the sensitivity analysis (Supplementary Table 1), the unadjusted model yielded a similar pattern of differences in FCR (p < 0.001, η2 = 0.220), whereas the unadjusted difference in social support was statistically significant but small in magnitude (p = 0.025, η2 = 0.021). These findings should be interpreted cautiously, These findings should be interpreted cautiously, as the small unadjusted difference in social support was no longer evident after adjustment for covariates.
Table 4.
Differences in FCR and social support across symptom phenotypes among breast cancer survivors (N = 357).
| Outcome | Emotion-reactive symptom phenotype (C1, n = 121) | Somatic-burden fatigue–pain phenotype (C2, n = 47) | Regulated low-symptom phenotype (C3, n = 189) | F | p | η2ₚ | Post-hoc (Tukey) |
|---|---|---|---|---|---|---|---|
| M ± SE | M ± SE | M ± SE | |||||
| FCR | 15.48 ± 0.68 | 15.19 ± 0.90 | 9.95 ± 0.62 | 42.986 | < 0.001 | 0.199 | C3 < C1, C2 |
| Social support | 39.13 ± 0.99 | 39.49 ± 1.32 | 40.99 ± 0.90 | 2.143 | 0.119 | 0.012 | n.s |
Values are adjusted means ± standard errors (SE) from ANCOVA models controlling for age, cancer stage, and medical burden; η2p = partial eta squared; n.s. = not significant.
Discussion
Using latent profile analysis, we explored symptom phenotypes among breast cancer survivors and examined differences in FCR and social support across these phenotypes. Three distinct symptom phenotypes were identified, namely, an Emotion-reactive symptom phenotype (33.9%), a Somatic-burden fatigue–pain phenotype (13.2%), and a Regulated low-symptom phenotype (52.9%), indicating that most breast cancer survivors experience a considerable psychological and physical symptom burden after diagnosis. After controlling for age, tumor stage, and medical burden, the three phenotypes showed significant differences in FCR levels, with an effect size in the medium-to-large range, whereas group differences in social support were no longer significant after adjustment for covariates.
The Emotion-reactive symptom phenotype, characterized by relatively high scores on Worrying and Feeling sad, accounted for 33.9% of all participants. Studies have shown that a psychological symptom cluster including anxiety, sadness, tension, distress, irritability, and difficulty concentrating is common among breast cancer patients receiving adjuvant chemotherapy30. In this study, 36.4% of patients were undergoing chemotherapy, which may partly explain the relatively high proportion of the Emotion-reactive symptom phenotype. In addition, the Somatic-burden fatigue–pain phenotype, characterized by the highest scores on Lack of energy, Difficulty sleeping, Pain, and Numbness/tingling, accounted for 13.2% of all participants. Notably, fatigue–sleep disturbance and psychological symptom clusters are the two most common symptom clusters among cancer patients31. However, unlike previous multi-symptom studies, the present study differentiated between two high-symptom phenotypes, namely, an “Emotion-reactive” and a “Somatic-burden fatigue-pain” phenotype, suggesting that psychological–emotional symptoms and physical symptoms do not always completely overlap. Some studies among breast cancer survivors have observed a similar pattern: one symptom cluster is dominated by anxiety, depression, and sleep disturbance, whereas another is characterized by fatigue, pain, and cognitive difficulties, and these clusters are associated with impairment in different functional domains32. The present study replicated this emotional–somatic differentiation pattern in Chinese breast cancer survivors, thereby supporting the use of symptom phenotypes as a useful tool for finely characterizing patient heterogeneity. It is also worth noting that time since diagnosis did not differ significantly across the three symptom phenotypes, suggesting that phenotype membership was not simply determined by the length of time since cancer diagnosis. Although symptom patterns may change across the treatment and survivorship trajectory, our findings indicate that both emotional- and somatic-dominant phenotypes may occur at different stages after diagnosis. These results suggest that symptom phenotypes should be interpreted by considering treatment exposure, perceived medical burden, functional status, and illness understanding.
In recent years, multiple studies exploring multi-symptom phenotypes in oncology populations have similarly identified a “low-symptom/recovered” phenotype and one or more “multidimensional high-symptom” phenotypes, with the high-symptom clusters being associated with markedly poorer functional status and health-related quality of life32. Additionally, symptom distress and illness uncertainty are positively associated with FCR, whereas social support exerts a protective effect17. Moreover, anxiety, depression, and sleep problems are positively and linearly associated with FCR levels33. In this study, analyses at the level of patient symptom phenotypes further confirmed this finding: patients with the Emotion-reactive and Somatic-burden fatigue–pain phenotypes had higher FCR levels, with effect sizes in the medium range. This may indicate that survivors with more pronounced psychological or somatic symptom burden are more likely to report recurrence-related threat perception and vigilance34,35. These findings are consistent with the hypothesized “multi-symptom burden, cognitive threat, and fear of cancer recurrence” pathway36, in which more severe or distressing symptoms may be interpreted by patients as possible signs of disease progression or recurrence. However, given the cross-sectional design, this pathway should be interpreted as a conceptual explanation. With respect to social support, the unadjusted analyses in this study showed that the low-symptom phenotype had the highest level of social support. However, this difference was no longer significant after controlling for age, cancer stage, and medical burden. This attenuation may be because social support is closely related to patients’ demographic and clinical context. Age may influence survivors’ family roles, social networks, and preferred sources of support, whereas cancer stage and perceived medical burden may affect both symptom experiences and the need for practical, emotional, and informational support. Therefore, the initially observed difference in social support across symptom phenotypes may partly reflect differences in age, cancer stage, and perceived medical burden37,38.
The occurrence of symptom clusters may not merely reflect a simple summation of individual symptoms. From a mechanistic perspective, the high levels of sleep disturbance, fatigue, pain, and numbness/tingling observed in the “Somatic-burden fatigue–pain” phenotype in this study may be interpreted in light of previously reported shared biobehavioral mechanisms, such as chronic inflammatory activation39, neuroendocrine dysregulation40, and sequence variations in genes encoding proinflammatory factors41. A recent study proposing a biobehavioral model42 suggested that systemic inflammation at the time of breast cancer diagnosis may concurrently drive tumor progression, symptom burden (such as fatigue, sleep disturbance, and emotional symptoms), and decline in quality of life. This model indicates that inflammatory and behavioral pathways may represent potential mechanisms underlying symptom clusters in breast cancer. Although these biological pathways were not measured in the present study, they provide a possible explanatory framework for the co-occurrence of these symptoms in the Somatic-burden fatigue–pain phenotype.
In addition, the Emotion-reactive symptom phenotype may be more characterized by worry and sadness, and its FCR level is comparable to that of the Somatic-burden fatigue–pain phenotype. Even when physical symptoms were relatively mild, patients experienced high levels of FCR when they lacked effective emotion regulation and adequate illness understanding. This finding is concordant with that of recent research emphasizing the cognitive and emotional mechanisms of FCR; thus, catastrophizing, hypervigilance, and negative beliefs play central roles in the development and maintenance of FCR43. Therefore, we recommend combining the interventions for this phenotype focus (i.e., cognitive-behavioral, mindfulness-based, or metacognitive psychological approaches) with appropriate provision of illness-related information and follow-up communication, to reduce unnecessary threat appraisals and excessive health care usage.
Notably, participants with the low-symptom phenotype accounted for half of the sample, indicating that a substantial proportion of breast cancer survivors can adapt and recover relatively well during follow-up, with a lighter symptom burden and comparatively low levels of FCR. This group tends to be older, have a relatively stable disease stage, a lighter medical burden, and preserved ADL. Routine follow-up combined with brief informational support is generally sufficient to meet their support needs. From a clinical practice perspective, combining symptom phenotypes with FCR can help rapidly identify high-risk patients and plan stratified intervention pathways. For example, patients with the Emotion-reactive symptom phenotype may be prioritized for psychoeducational interventions that focus on FCR management, whereas those with the Somatic-burden phenotype should receive comprehensive interventions that concurrently address pain, fatigue, and sleep disturbance. This stratified approach is highly consistent with the increasingly advocated symptom cluster-based precision nursing in international practice. In addition, emerging technologies may provide innovative tools to support phenotype-informed survivorship care. For example, digital symptom monitoring platforms, telehealth follow-up, and immersive technologies such as virtual reality or mixed reality may help provide tailored symptom education, emotional support, and self-management training for patients with different symptom phenotypes44,45. Such approaches may be particularly beneficial for patients with the Emotion-reactive phenotype, as they may benefit from technology-assisted psychological support and recurrence-related information. For patients with the Somatic-burden phenotype, these approaches may be more suitable for providing continuous symptom monitoring and remote management. However, the application of these technologies should also take into account potential limitations, including accessibility, digital literacy, cost, and the need to complement rather than replace direct clinical assessment and face-to-face care.
Limitations
Several limitations of this study must be acknowledged. First, the cross-sectional design does not allow causal inferences regarding the associations between symptom phenotypes, FCR, and social support, and it cannot capture transitions of phenotypes over time, even though previous studies have suggested that symptom cluster patterns may change during the course of treatment32. Second, although we included several important covariates, biological factors such as inflammatory markers and objective sleep indicators were not considered, which prevented direct testing of the biobehavioral model. Future studies could combine LPA with multiomics data or inflammatory biomarkers to investigate the pathophysiological pathways underlying different symptom phenotypes.
Conclusion
Three clinically meaningful symptom phenotypes were identified among Chinese breast cancer survivors; moreover, we found the Emotion-reactive and Somatic-burden phenotypes to be associated with significantly higher FCR levels, whereas the low-symptom phenotype was characterized by lower FCR levels and better symptom control. Adjustment for covariates revealed the differences in social support across phenotypes to be non-significant, suggesting that these effects may depend on contextual factors such as age, disease stage, and medical burden. Future research scope includes investigating the dynamic evolution of symptom phenotypes and their long-term impact on FCR and survival outcomes in longitudinal cohorts, as well as development and validation of targeted multimodal intervention programs based on high-risk phenotypes.
Supplementary Information
Below is the link to the electronic supplementary material.
Acknowledgements
Not application.
Author contributions
Han You and Weimin Liu were responsible for the conception and design of the work, as well as the acquisition and interpretation of data. Weimin Liu and Yali Su were conducted the data analysis and drafted the manuscript. Chong Chin Che, Ping Lei Chui, and Huaguo Zhang were responsible for supervision and critical review of the manuscript. Lihui Liu provided administrative and material support. All authors have read and approved the version to be published.
Funding
This study was not specifically funded by any public, commercial, or nonprofit sector institutions.
Data availability
The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.
Competing interests
All authors declare that there is no conflict of interest in the subject research and article writing process.
Ethical approval
This study was approved by the Ethics Committee of Beijing Shijitan Hospital, Capital Medical University (IIT2024-135-003), and all participants were fully informed about the study procedures and provided written informed consent. All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
These authors contributed equally: Han You and Weimin Liu.
References
- 1.Bray, F. et al. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J. Clin. 10.3322/caac.21834 (2024). [DOI] [PubMed] [Google Scholar]
- 2.Nolan, E., Lindeman, G. J. & Visvader, J. E. Deciphering breast cancer: from biology to the clinic. Cell186, 1708–1728. 10.1016/j.cell.2023.01.040 (2023). [DOI] [PubMed] [Google Scholar]
- 3.Duffy, S. W. et al. Mammography screening reduces rates of advanced and fatal breast cancers: Results in 549,091 women. Cancer126, 2971–2979. 10.1002/cncr.32859 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Nekhlyudov, L. et al. Developing a Quality of cancer survivorship care framework: Implications for clinical care, research, and policy. JNCI J. Natl. Cancer Inst.111, 1120–1130. 10.1093/jnci/djz089 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.He, Y. et al. Symptom burden, psychological distress, and symptom management status in hospitalized patients with advanced cancer: a multicenter study in China. ESMO Open7, 100595. 10.1016/j.esmoop.2022.100595 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Liu, W. et al. Effects of CBT-based interventions on health outcomes in breast cancer patients: a systematic review and Meta-analysis. Breast Cancer32, 689–704. 10.1007/s12282-025-01711-9 (2025). [DOI] [PubMed] [Google Scholar]
- 7.Puklin, L. S. et al. Improving lifestyle behaviors during chemotherapy for breast cancer: The Lifestyle, Exercise, and Nutrition Early After Diagnosis (LEANer) Trial. Cancer 10.1002/cncr.35280 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Ferreira, A. R. et al. Differential impact of endocrine therapy and chemotherapy on quality of life of breast cancer survivors: A prospective patient-reported outcomes analysis. Ann. Oncol.30, 1784–1795. 10.1093/annonc/mdz298 (2019). [DOI] [PubMed] [Google Scholar]
- 9.Carroll, J. E., Bower, J. E. & Ganz, P. A. Cancer-related accelerated ageing and biobehavioural modifiers: A framework for research and clinical care. Nat. Rev. Clin. Oncol.19, 173–187. 10.1038/s41571-021-00580-3 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Browall, M. et al. A prospective exploration of symptom burden clusters in women with breast cancer during chemotherapy treatment. Support Care Cancer25, 1423–1429. 10.1007/s00520-016-3527-1 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Langford, D. J. et al. Comparison of subgroups of breast cancer patients on pain and co-occurring symptoms following chemotherapy. Support Care Cancer24, 605–614. 10.1007/s00520-015-2819-1 (2016). [DOI] [PubMed] [Google Scholar]
- 12.Giesinger, J. M. et al. Thresholds for clinical importance were established to improve interpretation of the EORTC QLQ-C30 in clinical practice and research. J. Clin. Epidemiol.118, 1–8. 10.1016/j.jclinepi.2019.10.003 (2020). [DOI] [PubMed] [Google Scholar]
- 13.Bjerkeset, E., Röhrl, K. & Schou-Bredal, I. Symptom cluster of pain, fatigue, and psychological distress in breast cancer survivors: prevalence and characteristics. Breast Cancer Res. Treat180, 63–71. 10.1007/s10549-020-05522-8 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Jing, F. et al. Symptom profiles and related factors among breast cancer patients undergoing endocrine therapy: A latent profile analysis. Cancer Nurs.46, E297-e304. 10.1097/ncc.0000000000001125 (2023). [DOI] [PubMed] [Google Scholar]
- 15.Spurk, D., Hirschi, A., Wang, M., Valero, D. & Kauffeld, S. Latent profile analysis: A review and “how to” guide of its application within vocational behavior research. J. Vocat. Behav.120, 103445. 10.1016/j.jvb.2020.103445 (2020). [DOI] [Google Scholar]
- 16.Chen, F. et al. Fear of cancer recurrence and perceived pain in patients with breast cancer: A network analysis approach. Asia Pac. J. Oncol. Nurs.12, 100763. 10.1016/j.apjon.2025.100763 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Yu, Z., Sun, D. & Sun, J. Social support and fear of cancer recurrence among Chinese breast cancer survivors: The mediation role of illness uncertainty. Front Psychol.13, 864129. 10.3389/fpsyg.2022.864129 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Sebri, V. et al. Do you transfer your skills? From sports to health management in cancer patients. Front. Psychol.11, 2020. 10.3389/fpsyg.2020.00546 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Sebri, V. et al. Virtual reality for the promotion of interoception awareness and body image in breast cancer survivors: a study protocol. Front. Psychol.14, 2023. 10.3389/fpsyg.2023.1165905 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Tein, J. Y., Coxe, S. & Cham, H. Statistical power to detect the correct number of classes in latent profile analysis. Struct. Equ. Modeling20, 640–657. 10.1080/10705511.2013.824781 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Dziak, J. J., Lanza, S. T. & Tan, X. Effect size, statistical power and sample size requirements for the bootstrap likelihood ratio test in latent class analysis. Struct. Equ. Modeling21, 534–552. 10.1080/10705511.2014.919819 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Portenoy, R. K. et al. The Memorial Symptom Assessment Scale: an instrument for the evaluation of symptom prevalence, characteristics and distress. Eur. J. Cancer30a, 1326–1336. 10.1016/0959-8049(94)90182-1 (1994). [DOI] [PubMed] [Google Scholar]
- 23.Fu, L. et al. Validation of the simplified Chinese version of the memorial symptom assessment scale-short form among cancer patients. J. Pain Symptom Manage56, 113–121. 10.1016/j.jpainsymman.2018.03.024 (2018). [DOI] [PubMed] [Google Scholar]
- 24.Simard, S. & Savard, J. Screening and comorbidity of clinical levels of fear of cancer recurrence. J. Cancer Surv.9, 481–491. 10.1007/s11764-015-0424-4 (2015). [DOI] [PubMed] [Google Scholar]
- 25.Ting, S., Huaxia, L., Jing, T., Zhenzhen, D. & Zulin, W. Revision and reliability and validity testing of the Chinese version of the fear of cancer recurrence scale. Chin. J. Pract. Nurs.34, 1571–1576. 10.3760/cma.j.issn.1672-7088.2018.20.012 (2018). [DOI] [Google Scholar]
- 26.Xiao, S. in The Journal of Clinical Psychiatry Vol. 4 98 (1994).
- 27.Liu, Y. et al. Latent profile analysis of symptom clusters in children and adolescents with malignant tumors during radiotherapy. Cancer Nurs. 10.1097/ncc.0000000000001514 (2025). [DOI] [PubMed] [Google Scholar]
- 28.Nylund-Gibson, K. & Ten Choi, A. Y. frequently asked questions about latent class analysis. Transl. Issues Psychol. Sci.4, 440–461. 10.1037/tps0000176 (2018). [DOI] [Google Scholar]
- 29.Moore, E. W. G. & Quartiroli, A. Representing subpopulations with latent profile analysis: a non-technical introduction using exercisers’ goal orientation adoption profiles. J. Behav. Med. 10.1007/s10865-025-00596-5 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Qi, Y., Li, H., Guo, Y., Cao, Y. & Wong, C. L. Symptom clusters in breast cancer patients receiving adjuvant chemotherapy: A systematic review. J. Clin. Nurs.33, 4554–4567. 10.1111/jocn.17479 (2024). [DOI] [PubMed] [Google Scholar]
- 31.So, W. K. W. et al. Symptom clusters experienced by breast cancer patients at various treatment stages: A systematic review. Cancer Med.10, 2531–2565. 10.1002/cam4.3794 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Fardell, J. E., Tan, S. Y. C., Kerin-Ayres, K., Dhillon, H. M. & Vardy, J. L. Symptom clusters in survivorship and their impact on ability to work among cancer survivors. Cancers (Basel). 10.3390/cancers15215119 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Jiang, Y. et al. Association between fear of cancer recurrence and emotional distress in breast cancer: a latent profile and moderation analysis. Front. Psychiatry16, 1521555. 10.3389/fpsyt.2025.1521555 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Heathcote, L. C. et al. The bodily threat monitoring scale: Development and preliminary validation in adult and childhood cancer survivors. Psychooncology32, 1885–1894. 10.1002/pon.6236 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Guan, S. et al. Fear of cancer recurrence mediates the association between symptom burden and readiness for return to work in patients with lung cancer. Asia Pac. J. Oncol. Nurs.12, 100816. 10.1016/j.apjon.2025.100816 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Hu, Y. et al. A qualitative analysis of fear of cancer recurrence among Chinese nasopharyngeal cancer patients who are undergoing active treatment. Support Care Cancer33, 1136. 10.1007/s00520-025-10163-y (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Yu, X. et al. Impact of social support on body image during chemotherapy in patients with breast cancer: The chain mediating role of depression and self-efficacy. Asia Pac. J. Oncol. Nurs.12, 100664. 10.1016/j.apjon.2025.100664 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Li, H. et al. Dyadic effects of financial toxicity and social support on the fear of cancer recurrence in breast cancer patients and caregivers: An actor-partner interdependence mediation model. BMC Nurs.23, 378. 10.1186/s12912-024-02046-0 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.García-González, D., Medino-Muñoz, J., Romero-Elías, M., García-Foncillas, J. & Ruiz-Casado, A. Biological mechanisms of cancer-related fatigue in breast cancer survivors after treatment: A scoping review. J. Cancer Surv.19, 414–444. 10.1007/s11764-023-01477-z (2025). [DOI] [PubMed] [Google Scholar]
- 40.Bower, J. E. The role of neuro-immune interactions in cancer-related fatigue: Biobehavioral risk factors and mechanisms. Cancer125, 353–364. 10.1002/cncr.31790 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Pagliuca, M. et al. Long-term behavioral symptom clusters among survivors of early-stage breast cancer: Development and validation of a predictive model. JNCI J. Natl. Cancer Inst.117, 89–102. 10.1093/jnci/djae222 (2024). [DOI] [PubMed] [Google Scholar]
- 42.Di Meglio, A. et al. A bio-behavioral model of systemic inflammation at breast cancer diagnosis and fatigue of clinical importance 2 years later. Ann. Oncol.35, 1048–1060. 10.1016/j.annonc.2024.07.728 (2024). [DOI] [PubMed] [Google Scholar]
- 43.Omari, M. et al. Psychological distress and coping strategies in breast cancer patients under neoadjuvant therapy: A systematic review. Womens Health Lond.20, 17455057241276232. 10.1177/17455057241276232 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Ma, Z., Sun, L. & Zhang, Y. The role of virtual reality in breast cancer survivors: A scoping review. Prevent. Med. Rep.53, 103032. 10.1016/j.pmedr.2025.103032 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Jiménez-Díaz, A., Escrivá-Martínez, T., Herrero, R. & Baños, R. Evidence-based digital health interventions for breast cancer survivors’ care: An umbrella review. Int. Interv.41, 100846. 10.1016/j.invent.2025.100846 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.