Obesity is a major risk factor for onset of upper body breast cancer-related lymphedema (BCRL) [
Race is a social construct [
Identifying a profile for risk of greater BCRL severity is recognized as a great need in research [
The study population is drawn from 351 women enrolled in the ongoing Women in Steady Exercise Research (WISER) Survivor trial conducted within the University of Pennsylvania’s Transdisciplinary Research on Energetics and Cancer (TREC) Center. The WISER Survivor trial was approved by the University of Pennsylvania Institutional Review Board and all participants provided informed consent. WISER Survivor is a one-year randomized controlled weight loss and exercise intervention trial in sedentary overweight breast cancer survivors with breast cancer-related lymphedema (BCRL; see study
Eligibility requirements included (1) having stable (no cellulitis or flare-ups within the past 3 months) breast cancer-related lymphedema according to a standardized clinical evaluation by a Certified Lymphedema Therapist [
Interlimb arm volume difference was measured by perometry, operationalized as the difference between the left and right arms [
The demographic, physical, and clinical data were collected via self-report survey at study entry and included race, age, education level (nonoverlapping categories), employment status, total number of persons living in the household, presence of dependent children at home, years since breast cancer diagnosis, type of breast cancer treatment(s), and history of BCRL treatment. Education level cut-offs were based on previous examinations of education, self-rated health, and obesity, which use four categories (less than high school diploma, high school graduate, some college, and college graduate). However, due to sample size, the categories of less than high school and high school graduate were combined, resulting in a 3-category education variable [
Data analyses for this cross-sectional observational study were conducted using R version 3.2.2 (R Core Team (2015).
Table
Descriptive statistics.
( |
Mean/ |
SD/% | |
---|---|---|---|
Interlimb volume difference |
6.86 | 14.12 | |
Race | Black | 102 | 34.5% |
White | 194 | 65.5% | |
Body mass index (kg/m2) | 33.91 | 5.94 | |
Age at study entry (yrs) | 60 | 9 | |
Education | High school or less | 54 | 18.2% |
Some college | 102 | 34.5% | |
College grad or more | 140 | 47.3% | |
Retirement status | Not retired | 202 | 68.2% |
Retired | 108 | 32.4% | |
Hours per week of outside paid work | 25.88 | 18.4 | |
Number of people living in home | 3 | 1 | |
Dependent children at home | 117 | 39.5% | |
Cancer stage at diagnosis | 0 | 23 | 7.8% |
I | 90 | 30.4% | |
II | 90 | 30.4% | |
III | 81 | 27.4% | |
IV | 12 | 4.0% | |
Years since cancer treatment | 8.08 | 5.1 | |
Lymph nodes removed | Sentinel (<5 nodes) | 82 | 27.7% |
Axillary (≥5 nodes) | 214 | 72.3% | |
Adjuvant radiation therapy | 244 | 82.4% | |
Adjuvant chemotherapy | 246 | 83.1% | |
Adjuvant hormone therapy | 170 | 57.4% | |
Breast reconstruction | 118 | 39.9% | |
>25% lymphedema care adherence |
137 | 46.3% |
In Table
Multivariable linear regression for interlimb volume difference outcome.
Coefficient estimate |
Standard error |
| |
---|---|---|---|
Race (black ref.) | −0.26 | 1.79 | 0.89 |
Body mass index | 0.22 | 0.14 | 0.10 |
Age in years at study entry | 0.18 | 0.12 | 0.13 |
|
|
|
|
Education: college graduate versus high grad or less | −2.49 | 2.31 | 0.28 |
Retired | −0.88 | 2.41 | 0.72 |
Hours/week of paid outside work | −0.04 | 0.06 | 0.53 |
Number of people living in home | −0.76 | 0.66 | 0.25 |
Number of dependent children at home | 1.34 | 1.99 | 0.50 |
Cancer stage at diagnosis | 1.12 | 0.92 | 0.23 |
|
|
|
|
|
|
|
|
Radiation therapy | 0.04 | 2.37 | 0.99 |
Chemotherapy | −1.45 | 2.39 | 0.54 |
Hormone therapy | −0.44 | 1.69 | 0.79 |
Breast reconstruction | −0.23 | 1.89 | 0.91 |
|
|
|
|
Italicized items in table represent statistically significant covariates at
This analysis of breast cancer-related lymphedema (BCRL) in overweight Black and White women found no differences in interlimb arm volume by race or BMI but found large differences by education, suggesting a greater role for resource than race for BCRL severity. Interlimb volume difference is a measure of BCRL severity used by previous researchers and as part of clinical criteria [
The lack of difference in interlimb arm volume by race may mirror findings from a growing number of studies showing that the effects of race on BCRL onset are adjusted away in multivariable analysis [
Previous explorations of BCRL development have focused on the same clinical and physical risk factors that predispose certain populations to greater BCRL onset, such as node dissection type and BMI [
In the United States, educational attainment is considered to be a proxy measure for the larger construct of socioeconomic position, with low attainment being highly correlated with low income, high poverty risk [
Low-resource women may not have what they need to navigate worsening BCRL, especially in light of other competing demands that they face in day-to-day life. Still, in line with other studies with similar populations, adherence to BCRL care did not eliminate the interlimb volume disparity [
Contrary to some studies that have shown that BCRL can dissipate on its own in the first 18–24 months after treatment [
Behavioral interventions to reduce arm swelling have focused on the use of physical activity and weight loss to stabilize or lower BCRL severity [
While this study boasts a large community-based sample with broad coverage across a metropolitan area, results may not be generalizable, due to the use of a sample of participants who were healthy enough to enroll in a clinical trial. The study sample may be healthier or better-resourced than the overall population of cancer survivors. Nearly half of the study sample had a college education, compared to 29.2% of women over 35 in the US (US Census 2014 estimates), meaning that these women may be more resourced than their peers and that work and household indicators may not represent stressors for the relatively high-resource women who comprised the study population. Since it is not a population-based sample, determining whether or not differences in BCRL severity exist by race or other social factors in the general population is not possible; however, no existing population-based datasets of the general population include information on BCRL severity, making this study the best available sample to assess these relationships. Although perometry is a reliable tool for assessing upper body interlimb volume differences, the use of perometry for measuring interlimb knee [
This dataset was limited to a few proxy measures of socioeconomic position and resource level and did not have data on income or direct measures of psychosocial stress. Education is an imperfect proxy of resource level or socioeconomic position, which may explain the nonmonotonic relationship between education and BCRL severity that has been observed in other analyses of education, obesity, and health outcomes for population-based adult samples [
This study was designed to explore differences by race in interlimb arm volume differences among breast cancer survivors with breast cancer-related lymphedema (BCRL), accounting for social and physical risk factors for BCRL. In this sample of overweight women, neither race nor BMI was associated with greater interlimb volume, but the social factor of education level had the strongest relationship with interlimb differences. These results suggest a greater role of resource than race for BCRL severity and point to a possible approach to reduce disparities in the burden of BCRL. Interventions focused on reducing interlimb volume may be differentially useful across resource levels. This may stand in contrast to interventions focused on BCRL onset, which may still warrant specific focus on Black women. Future studies should further examine the roles of education level and socioeconomic differences with expanded and more precise measures of education and socioeconomic position.
The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.
The authors declare that there are no competing interests regarding the publication of this paper.
The authors thank and acknowledge the contributions of David Ebaugh, PT, Ph.D., Clinical Professor and Director of Human Anatomy (Health Sciences Department and Physical Therapy & Rehabilitation Sciences Department at the College of Nursing & Health Professions, Drexel University) and Bryan A. Spinelli, PT, Ph.D., OCS, CLT-LANA, Rehabilitation Clinical Specialist-Physical Therapy (Rhode Island Hospital) for their guidance on the measurement of the study’s primary outcomes. This study was supported by National Institute of Health Grant 1U54CA155850-01, with staffing support from the Leonard Davis Institute of Health Economics and the National Center for Research Resources and the National Center for Advancing Translational Sciences, National Institutes of Health, through Grant UL1TR000003. Dr. Dean is supported, in part, by the National Institutes of Health and National Cancer Institute Grant no. 1K01CA184288 and the Sidney Kimmel Cancer Center Core Grant P30CA006973.