This study represents a significant advancement in understanding lymphedema symptom patterns among BCS through latent class analysis (LCA). Four distinct latent classes were identified: “Severe Symptoms” group, “Movement-limitation and lymph-stasis” symptom group, “Lymph-stasis” symptom group, and “Low Symptoms” group. Factors such as ALND, radiotherapy, longer duration post-surgery, and lack of medical insurance were associated with decreased likelihood of belonging to the “Low symptoms” group. Network analysis revealed a decrease in symptom network density from severe to low symptom groups, with core symptoms including tenderness, firmness, arm-swelling, and heaviness. Prevalent symptoms among BCRL survivors aligned with previous findings, with swelling, heaviness, and tightness being most common.[10] The present study, targeting BCS with lymphedema, delineated specific symptom patterns, enhancing understanding for targeted interventions, such as stretch exercises for improving limb movements in the “Movement-limitation and lymph-stasis symptom group.”
Consistent with a previous study,[25] we observed a gradual increase in both the overall symptom severity and the average number of symptoms across classes 1 to 4. Despite common assumptions associating lymphedema symptoms with its severity, our findings showed no significant differences in lymphedema stage (Mild-to-moderate stage vs. Severe stage) among various symptom classes. This suggests that categorizing lymphedema patients solely based on symptom prevalence may not accurately reflect the progression of lymphedema. While assessing self-reported symptoms is widely accepted as a practical method for evaluating lymphedema, it is essential to recognize that symptoms are subjective experiences that include both the number of occurrences and the severity.
Our findings support the role of demographic and clinical factors in determining membership within latent symptom classes. Specifically, we observed that radiotherapy and ALND, recognized as risk factors for “Movement-limitation and lymph-stasis” symptom group (class 2) and “Lymph-stasis” symptom group (class 3), were associated with more invasive treatments that can lead to extensive tissue, nerve, vessel, and lymphatic damage, consequently exacerbating lymphedema symptoms. Additionally, participants in the first-year post-surgery were more likely to be in the “Low Symptom” group, possibly due to the predominant manifestation of BCRL within the initial two to three years following surgery.[26] Medical insurance emerged as a protective factor for the "Severe Symptom" group, possibly due to lower socioeconomic status hindering access to necessary rehabilitation and social support for lymphedema prevention and management. However, further demonstration is required as no similar reports were found.
Symptom networks provide a novel approach to understanding lymphedema symptomatology, allowing visualization and comprehension of microlevel interactions among lymphedema symptoms.[27] Network density is a predictive factor for identifying populations prone to comorbidities, and individuals with denser networks may exhibit lower treatment responsiveness.[28] Our study revealed a decreasing trend in network density from the “Severe symptom” group to the “Low symptom” group. This suggests that the “Severe Symptom” group may experience less favorable outcomes with standard symptom management strategies, highlighting the need for heightened attention and intensive interventions for survivors in this group. Results from network analysis identified tenderness, firmness, swelling, and heaviness as core symptoms for Class 1 to Class 4, respectively. These core symptoms exert the most significant influence on the entire network, making them potential targets for focused intervention.[29] Identifying high-risk populations within each lymphedema symptom subgroup allows efficient targeting for symptom management, aiding in the development of tailored interventions.[27] We recommend addressing upper limb symptoms alongside lymphedema treatments, and exploring effective strategies for specific core symptoms. For instance, some complementary therapies such as acupuncture or moxibustion have demonstrated efficacy in alleviating upper arm symptoms.[30, 31] Future research should focus on developing interventions to improve these specific core symptoms.
Strengths and limitations
While our study contributes novel insights into the dynamics of upper limb symptoms in BCS with lymphedema through the innovative use of latent class analysis (LCA) and network analysis, several limitations must be acknowledged. First, our study’s reliance on secondary data analysis constrained the inclusion of additional variables such as tumor stage or body mass index, potentially impacting the differentiation of latent classes. Additionally, restricted sample sizes within specific groups, such as non-insured individuals, may have compromised the statistical power of our analyses. Moreover, limitations within the original dataset hindered the conversion of interlimb circumference differences into limb volume, potentially affecting the accuracy of lymphedema diagnosis. Second, potential biases inherent in the parent studies from which the data were derived could have influenced the reliability and validity of our findings. The lack of control over the original data collection process introduces the possibility of biases or inconsistencies that could have affected subsequent analyses. These limitations underscore the need for further research with larger sample sizes and controlled designs to validate and build upon our findings.