Health-Related Quality of Life and Associated Factors in Long-Term Breast Cancer Survivors With and Without Lymphedema: A Cross-Sectional Study

Summary

This study compares health-related quality of life between long-term breast cancer survivors with and without lymphedema and examines clinical factors associated with poorer quality-of-life outcomes.

Abstract

Breast cancer-related lymphedema is a common long-term complication of breast cancer treatment that may continue to affect survivors well beyond active treatment. This study compared health-related quality of life between long-term breast cancer survivors with and without lymphedema and examined factors associated with poorer quality of life in a hospital-based cross-sectional study. Women who had undergone definitive breast cancer surgery at least 5 years earlier were recruited from outpatient and follow-up services at Shaoxing People's Hospital. Lymphedema status was determined using predefined criteria based on symptom duration and objective upper-limb measurements. Health-related quality of life was assessed using the Functional Assessment of Cancer Therapy-Breast (FACT-B), the EQ-5D-5L, and the EQ visual analogue scale (EQ-VAS). Clinical, treatment-related, and symptom-related variables were collected for adjusted analyses. Among 268 long-term breast cancer survivors, 69 were classified as having lymphedema and 199 as not having lymphedema. Participants with lymphedema had lower breast cancer-specific and generic health-related quality-of-life scores than those without lymphedema. In adjusted analyses, lymphedema status and pain remained independently associated with lower FACT-B total scores. These findings indicate that lymphedema remains associated with poorer health-related quality of life in long-term breast cancer survivors and support continued symptom-oriented surveillance during survivorship.

Introduction

Breast cancer remains the most frequently diagnosed malignancy in women, and continued gains in detection and treatment have enlarged the population living beyond the initial disease episode¹. As survival improves, the clinical focus necessarily extends beyond recurrence and mortality to the long-term conditions that shape daily life after treatment. Some of these conditions are apparent during active care, whereas others emerge later, persist over time, and become increasingly important as survivorship extends over years rather than months.

Breast cancer-related lymphedema belongs to the latter category. It is a chronic treatment-related condition that may develop shortly after surgery or remain clinically evident many years later, particularly in women exposed to more extensive axillary treatment or regional irradiation². Its reported frequency varies widely across studies, in part because case definitions differ and because different diagnostic approaches do not identify completely overlapping patient populations³. In this study, diagnostic classification combined symptom duration with standardized bilateral circumference and estimated limb-volume measurements, while alternative causes of arm swelling were excluded. Although this variability complicates comparisons across studies, it does not obscure the underlying clinical point that lymphedema remains one of the most persistent survivorship burdens after breast cancer treatment.

The condition is also not evenly distributed across survivors. Higher body mass index, axillary lymph node dissection, nodal irradiation, and multimodal treatment have repeatedly been associated with an increased risk of lymphedema⁴. Long-term follow-up data further suggest that this burden does not simply disappear with time; rather, a meaningful proportion of survivors continue to experience arm swelling or related symptoms several years after diagnosis⁵. The consequences extend beyond physical manifestations. Ongoing management costs, compression needs, clinic visits, work disruption, and treatment access can create broader practical, financial, and social burdens⁶.

This is clinically important because lymphedema is closely associated with health-related quality of life. Women with breast cancer-related lymphedema have been shown to report poorer quality of life than those without lymphedema, particularly in areas related to physical functioning, daily activities, and treatment-related discomfort⁷. That burden is often accompanied by fatigue, which may further impair functional recovery and overall health perception in survivorship populations already coping with upper-limb morbidity⁸. Among Chinese breast cancer survivors, lymphatic pain and related symptom burden have also received increasing attention, suggesting that the lived experience of lymphedema extends beyond visible swelling alone⁹.

Despite these findings, an important knowledge gap remains. Much of the literature has focused on incidence and risk prediction or on mixed survivorship populations without distinguishing long-term survivors from those who are relatively close to treatment. Earlier studies in Chinese women suggested that lymphedema after breast cancer treatment was associated with poorer quality of life, but evidence regarding long-term breast cancer survivors from hospital-based cross-sectional studies remains limited¹⁰. This gap is clinically relevant because long-term survivors often experience less acute treatment toxicity but continue to face chronic functional limitations, self-management demands, and cumulative survivorship burden.

The present study was therefore designed to compare health-related quality of life between long-term breast cancer survivors with and without breast cancer-related lymphedema in a hospital-based Chinese cross-sectional study and to examine the factors associated with quality-of-life impairment in this population.

Protocol

This single-center cross-sectional study was conducted at Shaoxing People’s Hospital through the breast surgery outpatient clinic, survivorship follow-up clinic, and hospital follow-up registry. The protocol was approved by the institutional ethics committee of Shaoxing People’s Hospital before recruitment began (approval no. JX20250291). All participants provided written informed consent before enrollment. The study was carried out in accordance with the Declaration of Helsinki and local regulations governing research involving human participants.

Study design and setting

The study was designed to compare health-related quality of life in long-term breast cancer survivors with and without breast cancer-related lymphedema and to examine factors associated with impaired quality of life. For this study, long-term survivorship was operationally defined as survival for 5 years or more after definitive breast cancer surgery. Recruitment was conducted from January 2023 through December 2024 at Shaoxing People’s Hospital. Eligible survivors were identified from outpatient schedules and the hospital follow-up database. All 268 participants completed a single in-person study visit, and no participant underwent a second clinical study visit. When clarification of a missing questionnaire response or verification of a chart variable was required, a supplementary telephone contact was permitted within 14 days; this administrative clarification did not involve repeat clinical assessment and did not constitute an additional study visit. Group classification was determined from the study-day clinical assessment together with chart review and was not redefined during the follow-up contact period. The study workflow is shown in Figure 1.

Figure 1. Study flow and final analytic grouping. (A) Flow diagram showing participant screening, eligibility assessment, reasons for exclusion, and inclusion in the final analytic cohort. A total of 296 participants were assessed for eligibility, 28 were excluded, and 268 participants were included in the final analysis. (B) Distribution of participants in the final analytic cohort, comprising 199 participants without breast cancer-related lymphedema (BCRL; 74.3%) and 69 participants with BCRL (25.7%). Group assignment was based on the predefined study criteria combining symptom duration and objective upper-limb assessment.Please click here to view a larger version of this figure.

Participant identification and eligibility screening

Participants were eligible if they were women aged 18–75 years, had pathologically confirmed unilateral primary breast cancer, had completed primary cancer treatment, had undergone definitive breast surgery at least 5 years earlier, had no evidence of local recurrence, distant metastasis, or second primary malignancy within the previous 6 months, were able to complete the questionnaires independently or with neutral reading assistance, and provided written informed consent. Participants were excluded if they had bilateral breast surgery, known primary lymphedema, major prior upper-limb trauma or orthopedic surgery affecting arm volume, acute cellulitis at the time of assessment, upper-extremity deep vein thrombosis, severe cardiac or renal edema, thyroid-related edema, or another medical condition likely to cause arm swelling independent of breast cancer treatment. Participants were also excluded if they had undergone a major upper-limb procedure within the previous 3 months or were unable to complete the core assessments because of severe cognitive or psychiatric impairment. Screening was completed in two stages. A research assistant first reviewed the electronic medical record for preliminary eligibility. Final confirmation was then made on site by a breast surgeon or a trained specialist nurse. When a discrepancy arose between chart review and direct assessment, the on-site clinical assessment was used for the final eligibility decision.

Sample size planning

The primary multivariable analysis used health-related quality of life as the dependent variable. The study prespecified 12 candidate predictors. A minimum of 15 evaluable participants per predictor was used as the planning rule, giving a minimum analytic sample of 180 participants. Allowing for incomplete questionnaires, post-screening exclusions, and missing covariate data, the recruitment target was set at a minimum of 210 participants. The final analysis set included participants who completed group classification and at least one core quality-of-life outcome.

Group classification for breast cancer-related lymphedema

Participants were assigned to either the lymphedema group or the non-lymphedema group using objective arm measurements, symptom duration, and the exclusion of alternative causes of swelling. Group assignment was completed before questionnaire scoring. The study protocol prespecified classification according to the presence or absence of BCRL. Objective measurements were used to confirm case status and were not intended to establish formal mild, moderate, or severe lymphedema stages.

Pre-measurement preparation

Participants rested in a seated position for 10 min before measurement. Vigorous upper-limb exercise was avoided during the preceding 12 h. Intensive edema-directed treatment, including manual lymphatic drainage, was not performed within 2 h before measurement. Participants who routinely used compression sleeves were asked to remove them at least 2 h before assessment, when clinically tolerable. If this was not possible, the reason was documented. A single study-day measurement did not overrule an established clinical history of persistent lymphedema documented in the medical record.

Circumference measurement procedure

Bilateral arm circumferences were measured by two trained research nurses using a non-stretch flexible measuring tape and recorded to the nearest 0.1 cm. Participants were assessed in a relaxed upright position with both arms resting naturally beside the trunk and the elbows gently extended. Circumference was measured at five predefined matched levels referenced to the olecranon: the wrist crease, 10 cm distal to the olecranon, the olecranon level, 10 cm proximal to the olecranon, and 15 cm proximal to the olecranon. Each level was measured twice. If the difference between the two readings was greater than 0.5 cm, a third measurement was obtained immediately. The final value was the mean of the two closest readings. Measurements were performed during fixed clinic time windows (08:30–11:30 or 14:00–16:30).

Diagnostic threshold for lymphedema

Breast cancer-related lymphedema was diagnosed after excluding acute infection, venous thrombosis, cancer recurrence, and systemic causes of edema. A participant was classified as having lymphedema when swelling-related symptoms had persisted for at least 3 months and either of the following objective criteria was present: a circumference difference of 2.0 cm or greater between the affected and contralateral arm at one or more predefined matched sites, or an estimated limb-volume increase of 10% or greater in the affected arm relative to the contralateral side¹¹. Limb volume was estimated from the five circumference values by constructing four truncated-cone segments and summing the segment volumes¹². Participants with a prior hospital diagnosis of breast cancer-related lymphedema remained in the lymphedema group if current symptoms or signs were still present, even when swelling was partially controlled at the study visit. The prespecified case definition required swelling-related symptoms to persist for at least 3 months. The protocol did not establish or retain a separate early-lymphedema analytic group for individuals with more recent swelling; therefore, the available dataset could not support an early-lymphedema subgroup comparison.

Symptom confirmation

After arm measurement, the assessor recorded heaviness, tightness, numbness, reduced range of motion, and difficulty related to wearing sleeves or bracelets or performing routine arm activities. Symptom onset and symptom duration were also recorded. For the descriptive analysis, an unweighted symptom-count score was calculated from five prespecified findings: heaviness, tightness, numbness, reduced range of motion, and difficulty wearing sleeves or bracelets or performing routine arm activities. Each finding was coded as absent (0) or present (1), giving a total score of 0–5. Symptom burden was categorized as low (0–1), moderate (2–3), or high (4–5). Participants with symptoms but without objective findings meeting the predefined diagnostic threshold were not classified as having lymphedema in the primary analysis. This symptom-count classification was used only for exploratory descriptive analysis and should not be interpreted as a standardized clinical measure of lymphedema severity.

Clinical and sociodemographic data collection

Clinical data were abstracted from the electronic medical record and verified with the participant when necessary. Variables included age, marital status, education, employment, insurance status, affected side, pathological stage, breast surgery type, axillary lymph node dissection, sentinel lymph node biopsy, chemotherapy, radiotherapy, endocrine therapy, targeted therapy, and time since surgery. Height and weight were measured on site. Height was recorded to the nearest 0.1 cm and weight to the nearest 0.1 kg. Body mass index was calculated as weight divided by height squared (kg/m²). Body mass index categories followed the Chinese adult classification: <18.5 kg/m² for underweight, 18.5–23.9 kg/m² for normal weight, 24.0–27.9 kg/m² for overweight, and ≥28.0 kg/m² for obesity¹³. Time since surgery was calculated from the date of definitive breast cancer surgery to the date of the study visit and recorded in months. If there was disagreement between participant recall and the medical record, the medical record was used. Detailed information on current lymphedema-directed treatment, including compression use, manual lymphatic drainage, rehabilitation participation, and treatment duration, was not systematically captured and was therefore not included as an analytic covariate.

Health-related quality-of-life assessment

Health-related quality of life was the primary study outcome. A breast cancer-specific instrument and a generic health-status instrument were administered in Chinese. Questionnaires were completed in a quiet room after the clinical assessment. Neutral reading assistance was provided when needed. Assessors were not permitted to comment on symptom severity, probable group assignment, or expected responses during questionnaire administration.

FACT-B

The primary breast cancer-specific outcome was the Functional Assessment of Cancer Therapy-Breast (FACT-B), Version 4. The scale contains 37 items, each scored from 0 to 4. Reverse-scored items were recoded according to the FACIT scoring manual¹⁴. The total score ranges from 0 to 148, with higher scores indicating better health-related quality of life. When more than 50% of items within a subscale were completed, the subscale score was prorated according to FACIT scoring guidelines. When completion fell below that threshold, the subscale score and any dependent total score were treated as missing. The validated simplified Chinese version of the FACT-B was administered in this study¹⁵.

EQ-5D-5L and EQ-VAS

Generic health status was assessed using the EQ-5D-5L, which includes five dimensions with five response levels each: mobility, self-care, usual activities, pain/discomfort, and anxiety/depression. Participants also completed the EQ visual analogue scale (EQ-VAS), scored from 0 to 100¹⁶. The EQ-5D-5L utility index was calculated using the Chinese value set¹⁷. If any EQ-5D-5L dimension was missing, the utility index was not calculated. Missing EQ-VAS responses were retained as missing only for that variable. The officially developed Chinese EQ-5D-5L descriptive system was used¹⁸.

Assessment of covariates potentially associated with quality of life

Additional variables were collected to capture symptom burden, psychological status, physical activity, and social support.

Pain intensity:

Pain in the affected arm, axilla, or ipsilateral chest wall during the previous 7 days was assessed using an 11-point numeric rating scale, ranging from 0 (no pain) to 10 (worst imaginable pain). For descriptive analyses, pain was categorized as mild (1–3), moderate (4–6), and severe (7–10), with a score of 0 reported separately.

Anxiety and depressive symptoms:

The Hospital Anxiety and Depression Scale (HADS) was used to assess anxiety and depressive symptoms¹⁹. Each subscale ranges from 0 to 21. Scores of 0–7 were classified as normal, 8–10 as borderline abnormal, and 11 or higher as indicating clinically relevant symptom burden. Continuous scores were used in the primary regression models. The validated Chinese version for cancer populations was administered²⁰.

Physical activity:

The International Physical Activity Questionnaire-Short Form (IPAQ-SF) was used to assess physical activity during the previous 7 days²¹. Total activity was calculated in metabolic equivalent (MET)-minutes/week using the standard scoring algorithm. Activity below 600 MET-min/week was classified as insufficient. Values of 600 MET-min/week or greater were classified as meeting the basic recommended level. Implausible responses were checked once with the participant. Unresolved inconsistencies were treated as missing. The Chinese short-form version has demonstrated acceptable reliability and validity²².

Social support:

Social support was assessed using the Social Support Rating Scale (SSRS)²³. Total scores range from 12 to 66, with higher scores indicating greater perceived support. For descriptive presentation, support was categorized as low (<23), moderate (23–44), and high (>44). Continuous scores were retained for regression analysis. The SSRS was originally developed for use in Chinese populations; therefore, a separate translated-version citation was not required.

Field implementation and quality control

All research nurses received standardized training before formal recruitment. Training covered participant approach, informed consent, anatomical positioning, arm-measurement technique, questionnaire administration, and data-entry procedures. A 10-participant pilot run was completed before formal enrollment, and pilot cases were excluded from the final dataset. Paper questionnaires were checked immediately after completion. Blank items were identified using a neutral prompt, and participants were allowed to decide whether to answer them. No directional prompting was permitted. All data were entered independently by two staff members. Automated and manual checks were then performed for score ranges, impossible values, treatment-history inconsistencies, abnormal circumference values, and date conflicts. Suspected errors were resolved by reviewing the original paper forms and medical records. If the correct value could not be confirmed, the original entry was retained and flagged. During the early recruitment phase, paired duplicate arm measurements were reviewed weekly to detect systematic measurement drift between assessors.

Statistical analysis

The analysis set included all participants with completed group classification and at least one core quality-of-life outcome. Normally distributed continuous variables were summarized as mean ± standard deviation and compared using the independent-samples t test. Non-normally distributed continuous variables were summarized as median and interquartile range and compared using the Mann–Whitney U test. Categorical variables were summarized as counts and percentages and compared using the Pearson chi-square test or Fisher’s exact test, as appropriate. The primary outcomes were the FACT-B total score, the EQ-5D-5L utility index, and the EQ-VAS score. Group differences between participants with and without lymphedema were examined first. Multivariable linear regression was then fitted with the FACT-B total score as the primary dependent variable. Prespecified candidate predictors were age, body mass index, time since surgery, axillary lymph node dissection, radiotherapy, chemotherapy, lymphedema status, pain score, HADS anxiety score, HADS depression score, physical activity, and social support. Variables with P < 0.10 in univariable analyses were eligible for entry into the multivariable model. Clinically important variables could be retained regardless of the univariable threshold. Multicollinearity was assessed before final model fitting, and a variance inflation factor greater than 5 was considered evidence of relevant collinearity. All tests were two-sided, and P < 0.05 was considered statistically significant. When the proportion of missing data for variables required in a specific model was less than 5%, complete-case analysis was used. When a key covariate had 5% or greater missingness, multiple imputation by chained equations was performed using 20 imputed datasets, and estimates were pooled using Rubin’s rules. Outcome variables were not imputed²⁴.

Data management and confidentiality

The participant identification file and the analytic dataset were stored separately. The identification file was password protected and accessible only to the principal investigator and the designated data manager. The analysis dataset contained study codes, group classification, measurement results, questionnaire scores, and de-identified clinical variables only. Electronic files were stored on a hospital server with controlled access and retained for at least 5 years after study completion. No identifiable data were released outside the institution without additional ethics approval.

Results

Study cohort and group classification

A total of 268 long-term breast cancer survivors were included in the final analysis after completion of group classification and availability of at least one core health-related quality-of-life outcome. Among them, 69 participants (25.7%) met the predefined criteria for breast cancer-related lymphedema, whereas 199 participants (74.3%) were classified as not having lymphedema. All 69 participants classified as having lymphedema met the predefined composite study criteria. A previously documented diagnosis was used only as supporting clinical history when current swelling was partially controlled and did not constitute a separate analytic classification group; therefore, participants were not divided into separate categories according to the route of classification. The study flow and participant exclusions are shown in Figure 1A, and the final analytic group distribution is shown in Figure 1B. The final cohort comprised two predefined comparison groups for subsequent analyses. There were no missing data for FACT-B total score, EQ-5D-5L utility index, or EQ-VAS (0/268, 0.0% for each outcome). Four participants (1.5%) had missing monthly household income data, and one participant (0.4%) had missing physical activity data. Because missingness was below 5%, the relevant adjusted analyses used complete-case data.

Clinical and sociodemographic characteristics

Baseline clinical and sociodemographic characteristics are presented in Table 1. The two groups were generally comparable with respect to age, education level, marital status, employment status, menopausal status, AJCC stage, surgery type, radiotherapy, chemotherapy, endocrine therapy, anti-HER2 therapy, number of comorbidities, and physical activity level. The mean age was 57.6 ± 9.1 years in the no-lymphedema group and 58.3 ± 8.1 years in the lymphedema group (P = 0.556). Several variables differed between groups. Participants with lymphedema had a longer interval since surgery than those without lymphedema (9.5 ± 3.1 vs. 8.5 ± 2.8 years, P = 0.017), a higher mean body mass index (25.1 ± 3.8 vs. 24.0 ± 3.4 kg/m², P = 0.033), and were more likely to have undergone axillary lymph node dissection (69.6% vs. 51.3%, P = 0.012). No statistically significant between-group differences were observed for the remaining baseline characteristics.

Variable	Category	No lymphedema (n = 199)	Lymphedema (n = 69)	Total (N = 268)	P value
Age, years	—	57.6 ± 9.1	58.3 ± 8.1	57.8 ± 8.8	0.556
Years since surgery	—	8.5 ± 2.8	9.5 ± 3.1	8.8 ± 2.9	0.017
Body mass index, kg/m²	—	24.0 ± 3.4	25.1 ± 3.8	24.3 ± 3.5	0.033
Education level	Junior high school or below	67 (33.7%)	20 (29.0%)	87 (32.5%)	0.77
	High school/technical secondary school	77 (38.7%)	29 (42.0%)	106 (39.6%)
	College or above	55 (27.6%)	20 (29.0%)	75 (28.0%)
Marital status	Married/cohabiting	172 (86.4%)	54 (78.3%)	226 (84.3%)	0.157
	Single/divorced/widowed	27 (13.6%)	15 (21.7%)	42 (15.7%)
Employment status	Retired	99 (49.7%)	33 (47.8%)	132 (49.3%)	0.511
	Employed	56 (28.1%)	24 (34.8%)	80 (29.9%)
	Unemployed/housework	44 (22.1%)	12 (17.4%)	56 (20.9%)
Monthly household income	<5,000 RMB	66 (33.2%)	22 (31.9%)	88 (32.8%)	0.531
	5,000–9,999 RMB	92 (46.2%)	35 (50.7%)	127 (47.4%)
	≥10,000 RMB	39 (19.6%)	10 (14.5%)	49 (18.3%)
	Missing	2 (1.0%)	2 (2.9%)	4 (1.5%)
Menopausal status	Premenopausal	53 (26.6%)	11 (15.9%)	64 (23.9%)	0.103
	Postmenopausal	146 (73.4%)	58 (84.1%)	204 (76.1%)
AJCC stage	I	59 (29.6%)	17 (24.6%)	76 (28.4%)	0.408
	II	91 (45.7%)	38 (55.1%)	129 (48.1%)
	III	49 (24.6%)	14 (20.3%)	63 (23.5%)
Surgery type	Mastectomy	142 (71.4%)	49 (71.0%)	191 (71.3%)	1
	Breast-conserving surgery	57 (28.6%)	20 (29.0%)	77 (28.7%)
Axillary lymph node dissection	No	97 (48.7%)	21 (30.4%)	118 (44.0%)	0.012
	Yes	102 (51.3%)	48 (69.6%)	150 (56.0%)
Radiotherapy	No	69 (34.7%)	16 (23.2%)	85 (31.7%)	0.106
	Yes	130 (65.3%)	53 (76.8%)	183 (68.3%)
Chemotherapy	No	44 (22.1%)	14 (20.3%)	58 (21.6%)	0.883
	Yes	155 (77.9%)	55 (79.7%)	210 (78.4%)
Endocrine therapy	No	38 (19.1%)	11 (15.9%)	49 (18.3%)	0.687
	Yes	161 (80.9%)	58 (84.1%)	219 (81.7%)
Anti-HER2 therapy	No	170 (85.4%)	60 (87.0%)	230 (85.8%)	0.91
	Yes	29 (14.6%)	9 (13.0%)	38 (14.2%)
Number of comorbidities	0	85 (42.7%)	34 (49.3%)	119 (44.4%)	0.671
	1	65 (32.7%)	18 (26.1%)	83 (31.0%)
	2	37 (18.6%)	14 (20.3%)	51 (19.0%)
	3	12 (6.0%)	3 (4.3%)	15 (5.6%)
Physical activity level	Low	65 (32.7%)	17 (24.6%)	82 (30.6%)	0.23
	Moderate	85 (42.7%)	33 (47.8%)	118 (44.0%)
	High	49 (24.6%)	18 (26.1%)	67 (25.0%)
	Missing	0 (0.0%)	1 (1.4%)	1 (0.4%)

Table 1: Baseline clinical and sociodemographic characteristics according to breast cancer-related lymphedema status.

Clinical and sociodemographic characteristics of long-term breast cancer survivors with and without breast cancer-related lymphedema (BCRL). Continuous variables are presented as mean ± standard deviation (SD). Categorical variables are presented as n (%). P values represent comparisons between participants with and without BCRL. Abbreviations: AJCC, American Joint Committee on Cancer; BMI, body mass index; BCRL, breast cancer-related lymphedema; HER2, human epidermal growth factor receptor 2; SD, standard deviation.

Main health-related quality-of-life outcomes

The primary health-related quality-of-life outcomes are summarized in Table 2. Across all three principal measures, participants with lymphedema had lower scores than those without lymphedema. For the breast cancer-specific instrument, the mean FACT-B total score was 104.89 ± 8.98 in the no-lymphedema group and 95.17 ± 8.38 in the lymphedema group, corresponding to a mean difference of 9.72 points (P < 0.001). For generic health status, the mean EQ-5D-5L utility index was 0.86 ± 0.07 in the no-lymphedema group and 0.79 ± 0.09 in the lymphedema group (P < 0.001). The mean EQ-VAS score was 77.37 ± 9.70 in the no-lymphedema group and 72.09 ± 8.75 in the lymphedema group (P < 0.001). The distributions of these outcomes are shown in Figure 2A–C. Across all three measures, the central tendency was lower in the lymphedema group, although the distributions partially overlapped.

Outcome	No lymphedema (n = 199)	Lymphedema (n = 69)	Mean difference (No − Yes)	P value
FACT-B physical well-being	21.40 ± 3.46	19.52 ± 3.63	1.87	<0.001
FACT-B social/family well-being	20.48 ± 3.39	19.36 ± 2.96	1.12	0.01
FACT-B emotional well-being	17.84 ± 3.61	16.42 ± 3.53	1.43	0.005
FACT-B functional well-being	18.50 ± 3.40	16.29 ± 3.59	2.21	<0.001
FACT-B additional breast cancer concerns	26.67 ± 4.20	23.57 ± 3.69	3.1	<0.001
FACT-B total score	104.89 ± 8.98	95.17 ± 8.38	9.72	<0.001
EQ-5D-5L utility index	0.86 ± 0.07	0.79 ± 0.09	0.07	<0.001
EQ visual analogue scale (EQ-VAS)	77.37 ± 9.70	72.09 ± 8.75	5.28	<0.001

Table 2: Health-related quality-of-life outcomes according to breast cancer-related lymphedema status.

Comparison of health-related quality-of-life outcomes between long-term breast cancer survivors with and without breast cancer-related lymphedema (BCRL). Continuous variables are presented as mean ± standard deviation (SD). Mean differences were calculated as the value for the no-lymphedema group minus the value for the lymphedema group. P values represent between-group comparisons. Abbreviations: BCRL, breast cancer-related lymphedema; EQ-5D-5L, EuroQol 5-Dimension 5-Level; EQ-VAS, EuroQol Visual Analogue Scale; FACT-B, Functional Assessment of Cancer Therapy–Breast; SD, standard deviation.

Figure 2. Health-related quality-of-life outcomes according to breast cancer-related lymphedema status. (A) Distribution of Functional Assessment of Cancer Therapy–Breast (FACT-B) total scores in participants with and without breast cancer-related lymphedema (BCRL). (B) Distribution of EuroQol 5-Dimension 5-Level (EQ-5D-5L) utility index scores in participants with and without BCRL. (C) Distribution of EuroQol Visual Analogue Scale (EQ-VAS) scores in participants with and without BCRL. Boxes represent the interquartile range, horizontal lines within the boxes represent the median, and whiskers extend to the most extreme observations within 1.5 times the interquartile range. Individual participant observations are overlaid as points. The mean and standard deviation (SD) for each group are reported below the corresponding plot. P values indicate between-group comparisons. Please click here to view a larger version of this figure.

FACT-B domain profile

Because the FACT-B total score differed between groups, domain-level results were examined. The results are summarized in Table 2, and the between-group profile is shown in Figure 3A,B. Mean scores for all five FACT-B domains were lower in the lymphedema group than in the no-lymphedema group. The largest absolute difference was observed for the additional breast cancer concerns domain, with mean scores of 26.67 ± 4.20 and 23.57 ± 3.69, respectively (P < 0.001). A marked between-group difference was also observed for functional well-being (18.50 ± 3.40 vs. 16.29 ± 3.59, P < 0.001). Smaller but statistically significant differences were observed for the physical well-being, social/family well-being, and emotional well-being domains.

Figure 3. Functional Assessment of Cancer Therapy–Breast domain scores according to breast cancer-related lymphedema status. (A) Comparison of the five Functional Assessment of Cancer Therapy–Breast (FACT-B) domain scores between participants with and without breast cancer-related lymphedema (BCRL), including physical well-being, social/family well-being, emotional well-being, functional well-being, and additional breast cancer concerns. Bars represent mean domain scores, and error bars represent standard deviations (SDs). P values indicate between-group comparisons for each domain. (B) Mean FACT-B domain-score profiles for participants with and without BCRL across the five FACT-B domains. Points represent group means, and connecting lines are provided to facilitate visual comparison. Higher scores indicate better health-related quality of life. Please click here to view a larger version of this figure.

Symptom-related score patterns

Symptom-related score patterns are presented in Figure 4A–C. Lower FACT-B total scores were observed among participants with greater symptom burden, although substantial overlap remained across individual observations. In Figure 4A, lower FACT-B total scores were observed with increasing lymphedema-related symptom burden. In Figure 4B, higher pain scores were associated with lower FACT-B total scores. In Figure 4C, higher HADS anxiety scores were also associated with lower FACT-B total scores in the unadjusted distribution.

Figure 4. Associations between symptom burden and Functional Assessment of Cancer Therapy–Breast total score. (A) Distribution of Functional Assessment of Cancer Therapy–Breast (FACT-B) total scores according to low, moderate, and high lymphedema-related symptom burden. Boxes represent the interquartile range, horizontal lines within the boxes represent the median, and whiskers extend to the most extreme observations within 1.5 times the interquartile range. Individual participant observations are overlaid as points. (B) Scatter plot showing the relationship between pain score and FACT-B total score. (C) Scatter plot showing the relationship between Hospital Anxiety and Depression Scale (HADS) anxiety score and FACT-B total score. In Panels B and C, each point represents one participant, solid lines represent fitted linear regression lines, and shaded bands represent the corresponding 95% confidence intervals. Please click here to view a larger version of this figure.

Multivariable linear regression for FACT-B total score

To assess whether the observed differences in health-related quality of life remained after adjustment for covariates, a multivariable linear regression model was fitted with FACT-B total score as the dependent variable. The results are presented in Table 3. After adjustment, lymphedema remained independently associated with a lower FACT-B total score (B = −8.33, 95% CI −10.89 to −5.77, P < 0.001). Pain score was also independently associated with a lower FACT-B total score (B = −1.67, 95% CI −2.34 to −1.00, P < 0.001). Age, body mass index, years since surgery, number of comorbidities, HADS anxiety score, and social support score were not significantly associated with FACT-B total score in the adjusted model. Retirement status showed a borderline association (B = −2.15, P = 0.079), although the 95% confidence interval included zero.

Variable	Regression coefficient (B)	Standard error (SE)	95% Confidence interval	P value
Intercept	117.538	6.55	104.636 to 130.439	<0.001
Employment status: Unemployed/housework vs. employed	−0.409	1.452	−3.269 to 2.452	0.779
Employment status: Retired vs. employed	−2.146	1.215	−4.539 to 0.248	0.079
Physical activity level: Low vs. moderate	−1.039	1.23	−3.461 to 1.383	0.399
Physical activity level: High vs. moderate	0.921	1.295	−1.630 to 3.472	0.478
Lymphedema status (yes vs. no)	−8.33	1.302	−10.894 to −5.765	<0.001
Age, years	−0.094	0.058	−0.209 to 0.022	0.111
Body mass index, kg/m²	−0.14	0.159	−0.453 to 0.173	0.378
Years since surgery	−0.139	0.186	−0.505 to 0.226	0.453
Number of comorbidities	−0.883	0.568	−2.001 to 0.235	0.121
Pain score (NRS 0–10)	−1.67	0.342	−2.344 to -0.996	<0.001
Hospital Anxiety and Depression Scale anxiety score	−0.357	0.469	−1.280 to 0.567	0.448
Social support score	0.058	0.054	−0.048 to 0.164	0.280

Table 3: Multivariable linear regression analysis of Functional Assessment of Cancer Therapy–Breast total score.

Multivariable linear regression analysis with Functional Assessment of Cancer Therapy–Breast (FACT-B) total score as the dependent variable. Regression coefficients (B), standard errors (SE), 95% confidence intervals (CIs), and P values are reported for variables retained in the final multivariable model. Abbreviations: B, regression coefficient; CI, confidence interval; FACT-B, Functional Assessment of Cancer Therapy–Breast; HADS, Hospital Anxiety and Depression Scale; NRS, numeric rating scale; SE, standard error.

Overall, participants with breast cancer-related lymphedema had poorer health-related quality-of-life outcomes than those without lymphedema across the primary outcome measures. After adjustment for potential confounders, lymphedema status and pain score remained independently associated with lower FACT-B total scores.

Data Availability:

The de-identified participant-level dataset and supporting materials underlying the findings of this study are publicly available in the Figshare repository at https://doi.org/10.6084/m9.figshare.33170417. Direct identifiers were removed before deposition, and the shared dataset contains only variables necessary to reproduce the reported analyses.

Discussion

The present study found that breast cancer-related lymphedema was associated with poorer health-related quality of life among long-term breast cancer survivors. This pattern was observed across the breast cancer-specific measure, the generic utility-based measure, and the self-rated health scale, rather than being limited to a single outcome instrument. Previous studies have likewise shown that the burden of lymphedema may persist for many years after breast cancer treatment, even after survivors have completed active treatment²⁵. Qualitative research has further suggested that this burden is often experienced as a continuing aspect of survivorship rather than as a discrete late event that resolves after cancer treatment²⁶. These findings support the use of multidimensional health-related quality-of-life assessment to characterize the long-term impact of breast cancer-related lymphedema.

The domain-level findings provide additional insight into how this burden is distributed across different aspects of health-related quality of life. The largest between-group differences were observed for additional breast cancer concerns and functional well-being, whereas statistically significant differences were also identified in the physical, emotional, and social/family well-being domains. This pattern suggests that the impact of breast cancer-related lymphedema extends beyond localized swelling and is reflected in broader aspects of daily functioning and survivorship. Earlier studies have likewise reported that arm heaviness, tightness, numbness, and related upper-limb symptoms are associated with poorer perceived physical health and reduced quality of life among breast cancer survivors²⁷. Recent clinical literature has also highlighted lymphatic pain as an important contributor to symptom burden, particularly when discomfort interferes with movement, routine activities, and perceived recovery²⁸. The adjusted and unadjusted findings were also clinically meaningful. The between-group difference of 9.72 points in FACT-B total score exceeded the established 7–8-point minimally important difference for this instrument, indicating that the observed impairment was not limited to statistical significance. The direction of the differences in FACT-B, EQ-5D-5L, and EQ-VAS was consistent with previous reports of poorer multidimensional health status among survivors with BCRL, although direct numerical comparisons remain limited by variation in lymphedema definitions, survivorship duration, and outcome instruments²⁹.

Several interacting biological and clinical pathways may explain this multidimensional impairment. Persistent failure of lymphatic drainage promotes interstitial fluid accumulation and may contribute to chronic tissue inflammation and fibrosis², increasing limb heaviness, tightness, pain, and restriction of movement. These symptoms can interfere with dressing, household activities, work, exercise, and sleep, while compression use, repeated self-management, fear of worsening swelling, and altered body image may add practical and psychological burden. The convergence of these physical and psychosocial pathways provides a plausible explanation for the lower physical, functional, emotional, social/family, and breast cancer-specific domain scores observed in participants with BCRL. Beyond the clinically meaningful FACT-B difference, the absolute between-group differences in the EQ-5D-5L utility index (0.07) and EQ-VAS (5.28 points) provide cross-instrument corroboration and indicate that the impairment extended from breast cancer-specific concerns to generic health status and self-rated health. In the adjusted health-related quality-of-life model, lymphedema status and pain score remained independently associated with lower FACT-B total score, whereas several demographic and treatment-related characteristics were not. This pattern suggests that current symptom burden may be more directly related to quality-of-life impairment than treatment history alone. Nevertheless, previous trial-based and meta-analytic evidence indicates that more extensive axillary treatment is associated with chronic upper-limb morbidity^30,31, and risk-modeling studies have identified body mass index, axillary lymph node dissection, radiotherapy, and the absence of rehabilitation planning as clinically relevant risk markers³². Prospective surveillance evidence further suggests that earlier detection and management may reduce progression to chronic arm lymphedema³³. Accordingly, established risk factors can help identify survivors who warrant continued monitoring, whereas pain, heaviness, tightness, functional limitation, and psychological distress should guide the urgency and intensity of rehabilitation. Survivors with persistent symptoms or declining quality-of-life scores should be referred promptly for lymphedema assessment, compression and exercise guidance, rehabilitation, pain management, and psychosocial support. This symptom-oriented approach may help prioritize rehabilitation resources and reduce the functional consequences of chronic BCRL.

Several limitations should be acknowledged. The cross-sectional design precludes assessment of temporal relationships, and the single-center setting may limit the generalizability of the findings to other survivorship populations. In addition, although the present study applied a structured clinical definition of lymphedema, currently available diagnostic approaches differ in sensitivity and may not identify symptomatic, subclinical, and chronic presentations equally well³⁴. Alternative approaches, including prospective longitudinal cohort studies and surveillance-based research, may further clarify the temporal development of lymphedema and its long-term impact on health-related quality of life. Within these limitations, the present findings support the value of structured survivorship surveillance and earlier symptom-oriented management, consistent with recent international guidance emphasizing prevention, monitoring, and risk-adapted care for breast cancer-related arm lymphedema³⁵. In practical terms, the findings suggest that breast cancer-related lymphedema in long-term survivors should be considered a chronic survivorship condition associated with pain, functional impairment, and reduced health-related quality of life rather than a localized physical sequela recognized only after obvious swelling develops.

In addition, the protocol did not establish a separate early-lymphedema group for survivors with swelling-related symptoms lasting less than 3 months, and complete data suitable for analysis of this stage were unavailable. Consequently, the present study could not evaluate health-related quality of life soon after lymphedema onset or determine how impairment changes with increasing disease duration. This limitation reduces the applicability of the findings to survivors with newly developed swelling. Future longitudinal studies should assess health-related quality of life at predefined intervals after lymphedema onset to identify when the greatest decline occurs and whether impairment subsequently stabilizes or continues to progress. Another limitation is that BCRL was analyzed as a binary exposure rather than according to standardized disease severity. Although bilateral circumference and estimated limb-volume measurements were obtained to confirm the diagnostic threshold, the study protocol did not prespecify severity categories, and the additional clinical information required for formal staging, such as tissue characteristics, pitting, reversibility, fibrosis, and longitudinal volume change, was not systematically collected. The symptom-burden categories shown in Figure 4A reflect the number of reported symptoms and should not be considered equivalent to objective lymphedema severity. Moreover, post hoc division of the 69 participants with BCRL into multiple severity categories could have produced small subgroups and potentially unstable multivariable estimates. Future adequately powered studies should prospectively apply a standardized severity-classification system and evaluate the independent association between increasing BCRL severity and health-related quality of life.

Detailed information on ongoing lymphedema-directed treatment and treatment duration was not systematically collected. Therefore, the study could not determine whether compression therapy, manual lymphatic drainage, exercise-based rehabilitation, or duration of treatment modified the association between BCRL and health-related quality of life. In addition, participants were recruited through hospital outpatient services, survivorship follow-up clinics, and the institutional follow-up registry. Survivors who remained engaged in hospital follow-up or who had persistent symptoms may therefore have been more likely to participate, whereas those with limited access to follow-up care or milder symptoms may have been underrepresented. This potential selection bias should be considered when generalizing the findings to the wider population of long-term breast cancer survivors.

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
EQ-5D-5L questionnaire with EQ-VAS, Chinese version	EuroQol Research Foundation	EQ-5D-5L, Chinese (China), paper self-complete version	Generic health-status instrument comprising the five-dimension descriptive system and the EQ visual analogue scale (EQ-VAS). The registered Chinese (China) version was administered, and utility index scores were calculated using the Chinese EQ-5D-5L value set.
FACIT scoring guidelines	FACIT.org / FACITtrans, LLC	FACT-B Version 4 scoring template, revised 11 June 2013	Used for reverse scoring, prorating eligible incomplete subscales, and calculating FACT-B domain and total scores according to the Version 4 scoring instructions.
FACT-B questionnaire, Chinese version	FACIT.org / FACITtrans, LLC	FACT-B, Version 4, Simplified Chinese	Breast cancer-specific health-related quality-of-life questionnaire containing 37 items. The Simplified Chinese Version 4 was used consistently for all participants.
Flexible circumference measuring tape	seca GmbH & Co. KG	seca 201	Non-stretch flexible circumference-measuring tape with millimeter graduation. Used for duplicate bilateral upper-limb circumference measurements at predefined matched anatomical levels.
Hospital Anxiety and Depression Scale (HADS), Chinese version	GL Assessment / Mapi Research Trust	HADS, Chinese version	Fourteen-item instrument comprising seven-item anxiety and depression subscales. The Chinese version was used consistently for all participants.
International Physical Activity Questionnaire-Short Form (IPAQ-SF), Chinese version	IPAQ Research Committee	IPAQ-SF, Chinese, short self-administered form, last 7 days; scoring protocol revised November 2005	Chinese short-form physical-activity questionnaire covering the previous 7 days. MET-min/week was calculated using the standardized IPAQ scoring protocol.
Portable stadiometer	seca GmbH & Co. KG	seca 213	Portable free-standing stadiometer used to measure standing height to the nearest 0.1 cm.
Social Support Rating Scale (SSRS), Chinese version	Xiao Shuiyuan, original developer	SSRS, 10-item Chinese scale, 1994	Chinese social-support instrument containing 10 items covering objective support, subjective support, and utilization of support. Scoring followed the original method described by Xiao Shuiyuan.
Spreadsheet software	Microsoft Corporation	Microsoft Excel for Microsoft 365	Used for independent double data entry, range checking, identification of inconsistencies, and data reconciliation before statistical analysis.
Stand-on digital floor scale	seca GmbH & Co. KG	seca 813	Digital flat scale with 200 kg capacity and 100 g graduation. Used to measure body weight to the nearest 0.1 kg.
Statistical analysis software	IBM Corporation	IBM SPSS Statistics, Version 29.0	Used for descriptive statistics, between-group comparisons, multivariable linear regression, multicollinearity assessment, complete-case analyses, and statistical data checks.