Skip to main content
NCBI home page
JAMA Network logoLink to JAMA Network
. 2026 Aug 18;9(8):e2629862. doi: 10.1001/jamanetworkopen.2026.29862

Long-Term Fatigue After Treatment for Breast Cancer

Nancy E Avis 1, Nathaniel S O’Connell 2, Shannon L Mihalko 1,3, Jennifer H Jordan 4,5, Amy C Ladd 4, Glenn J Lesser 6, Kristine C Olson 4, Kathryn E Weaver 1, Emily V Dressler 2, Pamela J Grizzard 4, Beverly Levine 1, Tonya L Moore 7, Bonnie Ky 8, Mary Helen Hackney 9, Peter Brubaker 3, Kerryn W Reding 10, Alexander R Lucas 11, Moriah P Bellissimo 4, F Gerard Moeller 12, Dalane W Kitzman 13,14, Craig A Hamilton 15, Heidi D Klepin 6, Joseph Yeboah 7, William Ntim 16, Vlad G Zaha 17, Ralph D’Agostino Jr 2, W Gregory Hundley 4,✉
PMCID: PMC13487558  PMID: 42611514

Key Points

Question

What mediates fatigue 24-months after breast cancer (BC) treatment?

Findings

In this cohort study of 381 patients with BC 2 years after receiving potentially cardiotoxic chemotherapy for BC, 34.6% of women experienced increased fatigue relative to their precancer treatment fatigue level. Independent of cardiovascular risk factors, including pretreatment and posttreatment left ventricular ejection fraction, only depressive symptoms and perceived stress were associated with and mediated the 24-month post-BC treatment fatigue measures.

Meaning

These findings suggest clinically relevant fatigue 2 years after initiating BC treatment is mediated partly by depressive symptoms and perceived stress irrespective of left ventricular ejection fraction.

This cohort study investigated cardiac, psychological, and cancer treatment factors associated with fatigue before and after breast cancer treatment.

Abstract

Importance

Patients with breast cancer (BC) may have long-term fatigue after treatment, which is associated with reduced quality of life.

Objective

To identify cardiac, psychological, and cancer treatment factors associated with fatigue before and 24 months after BC treatment.

Design, Setting, and Participants

This cohort study was conducted across multiple community hospital-based cancer centers among participants with stage I to III BC receiving potentially cardiotoxic chemotherapy or aromatase inhibitors and noncancer comparators. Participants were enrolled from May 2017 to July 2021 and followed up for 24 months. Data were analyzed between July 2023 and July 2026.

Exposures

Fatigue was assessed by the Functional Assessment of Chronic Illness Therapy Fatigue scale.

Main Outcomes and Measures

It was hypothesized that chemotherapy-associated changes in left ventricular function were associated with and mediated 2-year post-BC treatment fatigue. Outcome measures included fatigue, sociodemographic variables, hematocrit, left ventricular ejection fraction and circumferential strain, cardiovascular comorbidities, 6-minute walk distance, perceived stress, and depressive symptoms.

Results

From a total of 381 patients, patients with BC (236) and noncancer comparators (145) had a mean (SD) age of 54 (12) years. Two years after BC treatment, 33 (18.2%) receiving potentially cardiotoxic chemotherapy experienced clinically relevant fatigue (χ2 test P < .001 relative to aromatase inhibitors or controls), and 62 (34.6%) experienced a significant increase in fatigue relative to their precancer treatment fatigue level (χ2 test P < .001, relative to aromatase inhibitors and controls). Independent of pretreatment and posttreatment left ventricular ejection fraction, diabetes status, hypertension, tobacco use, age, body mass index, hematocrit, receipt of potentially cardioprotective medications, and chemotherapy regimens, only depressive symptoms (estimate, −0.77; 95% CI, −0.83 to −0.71; P < .001) and perceived stress (−0.32; 95% CI, −0.41 to −0.24; P < .001) were associated with and mediated 24-month post-BC treatment fatigue measures. Baseline adjusted changes in 6-minute walk distance, left ventricular ejection fraction, and left ventricular strain were not associated with 24-month post-BC treatment fatigue levels.

Conclusions and Relevance

In this cohort study of patients with BC, clinically relevant fatigue increased in 34% of patients 2 years after initiating cancer treatment mediated partly by depressive symptoms and perceived stress regardless of cardiovascular risk factors, age, or change in left ventricular ejection fraction and/or strain.

Trial Registration

ClinicalTrials.gov Identifier: NCT02791581

Introduction

Post–breast cancer (BC) treatment prevalence estimates of fatigue range from 20% to 62% depending on the population, type of treatment, time since treatment, and method of assessment.1,2,3,4 It is unknown whether the interrelationship of preexisting or dynamic cancer therapy–associated change in cardiovascular (CV), physiologic, psychological, or metabolic factors mediate BC treatment–related fatigue.

We hypothesized that cancer chemotherapy associated with changes in left ventricular (LV) function (known to occur early during cancer treatment) were associated and mediated 2-year post–BC treatment fatigue. To address this hypothesis, we conducted a multisite prospective cohort study of women treated with adjuvant chemotherapy and/or an aromatase inhibitor for BC—including a noncancer comparison cohort5,6— to determine the pretreatment to 24-month posttreatment prevalence of fatigue accounting for age, race and ethnicity, cancer therapy, LV function, and sociodemographic, psychological, physiologic, and CV risk factors.

Methods

Understanding and Predicting Breast Cancer Events after Treatment (UPBEAT, WF-97415, funded via R01CA199167) is a multisite prospective cohort study of women diagnosed with stage I to III BC7 and noncancer comparators operating through the Wake Forest National Cancer Institute Community Oncology Research Program Research Base (UG1CA189824) and the Eastern Cooperative Oncology Group-American College of Radiology Imaging Network Cancer Research Group (UG1CA189828) under a National Cancer Institute Central institutional review board–approved protocol and registered with ClinicalTrials.gov (NCT02791581). All participants provided witnessed written informed consent. Participants were enrolled across 66 community hospitals, where up to 85% of women receive BC treatment.8,9 Participants were registered between August 2017 and July 2021 and followed up for 24 months after their baseline or pretreatment assessment. The study design and rationale along with detailed inclusion and exclusion criteria have been described.2,10 This study is reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline for cohort studies (eTable 8 in Supplement 1).

Study Population

All participants, including noncancer comparators, were at least 18 years old. Eligibility included inflammatory, newly diagnosed, or locally recurrent BC being treated with curative intent scheduled to receive chemotherapy and/or aromatase inhibitor (AI) with an Eastern Cooperative Oncology Group Performance Status between 0 and 2.11 Exclusion criteria included a contraindication to CV magnetic resonance imaging, an LV ejection fraction of less than 50% if assessed prior to treatment, current pregnancy or lactation, uncontrolled hypertension, or a preexisting inflammatory disease (eg, systemic lupus). A convenience sample of female family and friends without a history of cancer, breast surgery, or chemotherapy served as a comparator group.

Persons were categorized as (1) receiving chemotherapy inclusive of doxorubicin along with other chemotherapy agents or an AI (chemotherapy inclusive of doxorubicin), (2) receiving chemotherapy possibly with an AI but without doxorubicin (eg, paclitaxel, cyclophosphamide, capecitabine, gemcitabine) (chemotherapy without doxorubicin), or (3) receiving an AI only (anastrozole, letrozole, or exemestane) and no doxorubicin or other chemotherapy (AI, no chemotherapy) (eTable 1 in Supplement 1). For each participant, receipt of radiation treatment for breast cancer was recorded. Participant’s age of less than or at least 52 years was used respectively to characterize premenopausal or postmenopausal status.12,13

Participant characteristics including age, race, height, weight, body mass index, brachial artery systolic and diastolic blood pressure (collected according to American Heart Association criteria),14 heart rate, and receipt of cardioactive medications (beta blockers, angiotensin receptor antagonists, angiotensin converting enzyme inhibitors, calcium antagonists, and diuretics) were collected at each study visit. Race and ethnicity were self-reported at enrollment using fixed case report form categories consistent with the National Institutes of Health categories.

Study Procedures

In participants with BC, complete data inclusive of magnetic resonance imaging were collected pretreatment (postdiagnosis, before postsurgical or nonsurgical treatment) and then 3 and 24 months later. Comparators contributed magnetic resonance imaging and laboratory measures at enrollment and 24 months, and other data at enrollment and at 3 and 24 months.

Outcome

Fatigue was measured by the Functional Assessment of Chronic Illness Therapy fatigue scale15 across all study visits.16 Scores range from 0 to 52, with higher scores indicating lower fatigue levels.16 A fatigue score of less than or equal to 3017 and a change in score greater than or equal to 5 points was regarded as clinically significant.18

Risk Factors

Factors of fatigue included sociodemographics, CV risk factors, cancer-related factors, laboratory and imaging biomarkers of subclinical CV disease, and psychological factors.7 Medical treatment was retrieved from participants’ medical records. The comorbidities diabetes, hypertension, and a history of coronary artery disease were self-reported (yes, no, or do not know to ever receiving a diagnosis) as was smoking history (never, former, or current). Laboratory data pertaining to glucose, serum creatinine, C-reactive protein, and cholesterol status were collected.

Cardiovascular magnetic resonance assessments of LV volumes, mass, ejection fraction, and global circumferential strain were performed with a standard set of acquisition parameters across a range of 1.5 and 3.0 Tesla scanners from General Electric, Philips, and Siemens according to previously published methods.19,20 Circumferential strain and other LV measures were obtained using cine white blood steady state free precession techniques with a 256 × 128 matrix, a 40-cm field of view, a 10-millisecond repetition time, a 4-millisecond echo time, a 20-degree flip angle, an 8-mm–thick slice with a 2-mm gap, and a 40-millisecond temporal resolution.19,20 All images were analyzed (Medis Medical Imaging, Leiden, Netherlands) by readers in a central core laboratory blinded to all identifiers as an unpaired, blinded read. Submaximal exercise capacity was measured by the 6-minute walk distance (6MWD), an independent factor of CV mortality.21 Psychological factors included depressive symptoms (measured by the 10-item Center for Epidemiological Studies Depression scale [CESD] in which scores range from 0 to 30 with a score of 10 or greater indicating significant depression),22 and perceived stress (measured by the Cohen 4-item Perceived Stress Scale [PSS]).23

Statistical Analysis

Expanded statistical methods are provided in the eMethods in Supplement 1. All analyses were conducted between October 2024 through April 2025 using R version 4.1.1 (R Project for Statistical Computing).24 Continuous variables were summarized as mean (SD), and categorical variables as counts and percentages. Analyses followed an intention-to-treat framework based on treatment group at enrollment. Missing longitudinal data were addressed using multiple imputation by chained equations, with estimates via Rubin rules.25 Sensitivity analyses compared multiply imputed and complete-case results.

To evaluate differences in fatigue at 24 months between treatment groups and comparators, linear regression models were fit with and without adjustment for pretreatment fatigue (ie, baseline-adjusted analysis of covariance), which provides greater power and precision than change-score or mixed-effects analyses for pre-post data.26,27 Associations between candidate factors and 24-month fatigue were explored using separate pretreatment fatigue adjusted regression models. To identify variables independently associated with fatigue, least absolute shrinkage and selection operator (LASSO) regression was used for variable selection from prespecified demographic, clinical, imaging, and psychological factors28,29 and selected variables were then included in multivariable models.

Potential mediating pathways linking treatment, depressive symptoms, perceived stress, and fatigue over time were evaluated using structural equation models with latent change scores.30,31,32,33 All statistical tests were 2-sided and tested at a P = .05 level of significance.

Results

This cohort of 381 encompassed patients with BC (236) and noncancer comparators (145) who had a mean (SD) age of 54 (12) years, and reported 6 Asian (1.6%), 24 (16.6%) Black, and 110 (75.9%) White individuals (Table 1; eTable 2 in Supplement 1). Although 403 participants enrolled in the study (258 participants with BC and 145 comparators), 22 participants withdrew due to feeling overwhelmed prior to any participation. Sixty-one of 381 participants (16%) dropped out (Figure 1) before 24-month; 35 (57.4%) were comparators, and dropouts were younger (50.7 vs 54.4 years, t test P = .03). Overall, there was not a significant difference in pretreatment fatigue between those who dropped out vs those who remained in the study.

Table 1. Participant Characteristics by Cancer Treatment Group.

Characteristics Participants, No. (%) Standardized mean difference
Overall (N = 381) Inclusive of doxorubicin (n = 117) Chemotherapy, without doxorubicin (n = 84) Aromatase inhibitors, no chemotherapy (n = 35) Comparators without cancer (n = 145) All 4 groups Cancer groups only
Age, mean (SD), y 53.8 (12.4) 52.9 (11.0) 56.8 (10.8) 62.5 (7.1) 50.7 (13.9) 0.62 0.67
Age groups, y
<52 150 (39.4) 52 (44.4) 25 (29.8) 2 (5.7) 71 (49.0) 0.6 0.66
≥52 231 (60.6) 65 (55.6) 59 (70.2) 33 (94.3) 74 (51.0)
Race
American Indian Alaska Native 0 0 0 0 0 0.52 0.45
Asian 6 (1.6) 3 (2.6) 0 0 3 (2.1) NA NA
Black 66 (17.3) 23 (19.7) 14 (16.7) 5 (14.3) 24 (16.6) NA NA
Native Hawaiian Pacific Islander 2 (0.5) 0 1 (1.2) 0 1 (0.7) NA NA
Not reported 8 (2.1) 0 3 (3.6) 2 (5.7) 3 (2.1) NA NA
White 289 (75.9) 87 (74.4) 66 (78.6) 26 (74.3) 110 (75.9) NA NA
Unknown 10 (2.6) 4 (3.4) 0 (0) 2 (5.7) 4 (2.8) NA NA
Height, mean (SD), m 1.65 (0.07) 1.65 (0.06) 1.65 (0.07) 1.65 (0.06) 1.66 (0.07) 0.06 0.00
Weight, mean (SD), kg 78.7 (18.3) 80.1 (20.3) 78.6 (16.4) 80.0 (16.1) 77.1 (18.3) 0.1 0.06
Body mass index, mean (SD)a 28.7 (6.4) 29.2 (6.8) 28.8 (5.7) 29.3 (5.2) 28.1 (6.7) 0.1 0.06
Systolic BP, mean (SD), mm Hg 122 (16.4) 121 (16.4) 127 (19.2) 129 (13.7) 118 (13.9) 0.41 0.33
Diastolic BP, mean (SD), mm Hg 68.7 (12.0) 67.7 (12.6) 70.7 (12.8) 69.7 (11.4) 68.2 (11.2) 0.14 0.16
Hypertension
Yes 109 (28.6) 31 (26.5) 33 (39.3) 17 (48.6) 28 (19.3) 0.43 0.38
Do not know 3 (0.8) 0 2 (2.4) 1 (2.9) 0 NA NA
Diabetes
Yes 31 (8.1) 10 (8.5) 12 (14.3) 2 (5.7) 7 (4.8) 0.29 0.28
Do not know 4 (1.0) 1 (0.9) 1 (1.2) 2 (5.7) 0 NA NA
Smoking status
Never 229 (60.1) 68 (58.1) 54 (64.3) 24 (68.6) 83 (57.2) 0.29 0.23
Former 69 (18.1) 28 (23.9) 14 (16.7) 8 (22.9) 19 (13.1) NA NA
Current 10 (2.6) 7 (6.0) 1 (1.2) 2 (5.7) 0 NA NA
History of CAD
Yes 7 (1.8) 1 (0.9) 2 (2.4) 2 (5.7) 2 (1.4) 0.28 0.32
Do not know 4 (1.0) 0 2 (2.4) 2 (5.7) 0 NA NA
CV medicationsb
Yes 62 (16.3) 15 (12.8) 18 (21.4) 13 (37.1) 16 (11.0) 0.36 0.39
Heart rate, mean (SD), beats/min 73.4 (11.8) 75.0 (12.2) 72.8 (12.8) 71.9 (13.0) 72.6 (10.3) 0.13 0.16
Hematocrit, mean (SD), % 39.6 (3.2) 38.7 (3.6) 39.8 (3.1) 39.6 (2.3) 40.2 (3.0) 0.24 0.23
Creatinine, mean (SD), mg/dL 0.781 (0.137) 0.753 (0.138) 0.767 (0.121) 0.843 (0.180) 0.800 (0.128) 0.34 0.39
Cancer stage
No cancer 145 (38.1) 0 0 0 145 (100) 1.2 1.2
I 87 (22.8) 19 (16.2) 38 (45.2) 30 (85.7) 0 NA NA
II 122 (32.0) 81 (69.2) 36 (42.9) 5 (14.3) 0 NA NA
III 27 (7.1) 17 (14.5) 10 (11.9) 0 0 NA NA
Radiation 158 (41.5) 88 (75.2) 61 (72.6) 9 (25.7) 0 0.66 0.75
Depression baseline, mean (SD) 5.30 (4.66) 6.42 (4.72) 5.84 (4.72) 5.18 (4.98) 4.07 (4.23) 0.28 0.17
PSS baseline, mean (SD) 3.95 (2.94) 4.53 (3.08) 4.20 (3.03) 3.03 (2.59) 3.54 (2.75) 0.3 0.35
Severe fatigue at baseline, No./total No. (%) 34/371 (9.2) 16/115 (13.9) 10/83 (12.0) 5/34 (14.7) 3/139 (2.2) NA NA
Severe fatigue at 24 mo, No./total No. (%) 38/320 (11.9) 19/102 (18.6) 14/79 (17.7) 2/30 (6.7) 3/109 (2.8) NA NA
Significant decline from baseline to 24 mo, No./total No. (%) 91/318 (28.6) 33/101 (32.7) 29/78 (37.5) 8/30 (26.7) 21/109 (19.3) NA NA
Open in a new tab

Abbreviations: BP, blood pressure; CAD, coronary artery disease; CV, cardiovascular; NA, not applicable; PSS, perceived stress scale.

a

Body mass index calculated as weight in kilograms divided by height in meters squared.

b

CV medications include angiotensin-converting enzyme inhibitors, beta blockers, calcium channel blockers, and diuretics.

Figure 1. Matrix of Number of Participants at Each Stage of the Study.

Table of participant counts by visit time point. The table has four columns and six rows, including the two header rows. The first header row contains a merged header centered over the first three columns reading Visit; the fourth column header reads Participants, No. The second header row lists the first three column headers as Pretreatment, 3-mo, and 24-mo, with the fourth column continuing as the participant count column. Four data rows follow, each using filled circular markers to indicate which visits occurred for that row. Data row 1: Pretreatment contains a single filled circle; 3-mo is blank; 24-mo is blank; Participants, No. is 14. Data row 2: Pretreatment contains a filled circle; 3-mo contains a filled circle; 24-mo is blank; Participants, No. is 44. Data row 3: Pretreatment contains a filled circle; 3-mo is blank; 24-mo contains a filled circle; Participants, No. is 3. Data row 4: Pretreatment contains a filled circle; 3-mo contains a filled circle; 24-mo contains a filled circle; Participants, No. is 320. The final row contains blanks in the first three columns and the text Total No., 381 in the Participants, No. column.

Open in a new tab

Approximately 84% of all participants provided data at all 3 visits used in the study.

Univariable Associations

Table 1 summarizes the prevalence of clinically significant fatigue and changes in fatigue over 24 months. Figure 2 displays the association between cancer treatment groups relative to comparators in pretreatment and 24-month fatigue. Relative to comparators, participants in all 3 BC groups reported worse (ie, lower) pretreatment and 24-month posttreatment fatigue. Compared with those without cancer, participants receiving chemotherapy inclusive of doxorubicin and chemotherapy without doxorubicin continued to report worse fatigue scores over time (mean differences, −3.93 [95% CI, −6.08 to −1.79]; P < .001; and −4.37 [95% CI, −6.81 to −1.94; P < .001], respectively). However, fatigue scores for the AI, no chemotherapy group were not different from the comparators (mean difference, −2.46 [95% CI, −5.65 to 0.73; P = .13]). Notably, pretreatment fatigue was associated with 24-month fatigue (β, 0.35 [95% CI, 0.25 to 0.46]; P < .001), indicating that worse fatigue at pretreatment was associated with worse fatigue 24 months later.

Figure 2. Line Chart of Fatigue Over Time by Group.

Line chart of fatigue score versus time in months for four treatment groups. Single-panel line chart with error bars and colored circular markers. Vertical axis label Fatigue score, with tick marks at 25, 30, 35, 40, 45, and 50. Horizontal axis label Time, mo, with labeled ticks at 0, 3, and 24. Four series correspond to the legend titled Group in the lower right: orange Comparators without cancer; bright blue Chemotherapy inclusive of doxorubicin; dark teal Chemotherapy without doxorubicin; light gray Aromatase inhibitors, no doxorubicin. Each time point has a mean marker with vertical 95 percent confidence interval whiskers, and markers within each group are connected by thin black lines. At time 0, the orange mean is near 46 to 47, with a second orange point near 45 to 46 at 3 months, and another orange point near 45 to 46 at 24 months. At time 0, the bright blue and dark teal means are near 42 to 43, then both drop at 3 months to roughly 31 to 33, then rise by 24 months to about 40 to 41. The light gray series is near 40 to 41 at time 0, near 40 at 3 months, and near 41 to 42 at 24 months. Significance brackets appear near the left around baseline with text P less than .001, and near the right around 24 months with one bracket labeled P equals .11 and another labeled P less than .001, each bracket spanning between specific group markers at that time.

Open in a new tab

Unadjusted fatigue scores (y-axis) over time in months (x-axis) are shown for the following groups: orange indicates comparators without cancer; light blue indicates chemotherapy, inclusive of doxorubicin; dark blue indicates chemotherapy, without doxorubicin; and gray indicates aromatase inhibitors, no chemotherapy. Mean with 95% CIs are shown. Scores range from 0 to 52, with higher scores indicating lower fatigue levels.

eTable 3 in Supplement 1 displays participant biomarker, LV function, and key construct scores (fatigue, PSS, CESD) by treatment groups. eTable 4 in Supplement 1 reports associations between participant pretreatment characteristics and their 24-month changes with the 24-month fatigue, unadjusted and adjusted for pretreatment fatigue. When adjusted for a participant’s pretreatment fatigue, we observed univariable associations between higher pretreatment PSS (P = .002), 24-month increase in PSS (P = .01), and 24-month worsening of depressive symptoms (P < .001). Baseline-adjusted changes in hematocrit, 6MWD, left ventricular ejection fraction (LVEF), and LV circumferential strain were not associated with 24-month fatigue.

Multivariable Associations

Table 2 presents results (all participants included) from a multivariable model with variables selected via LASSO regression (eTable 6 in Supplement 1). Pretreatment fatigue and 3-month fatigue were associated with 24-month fatigue. Higher CESD and PSS at 24 months were also each independently associated with 24-month fatigue scores. Notably, cancer treatment was not among the selected variables.

Table 2. Reduced Multivariable Regression Models Selected by LASSO Regressiona.

Term Estimate (95% CI) P value
Multivariable model of 24-mo fatigue
Intercept 34.75 (33.44 to 36.07) <.001
Pretreatment fatigue 0.15 (0.12 to 0.18) <.001
3-mo Fatigue 0.16 (0.14 to 0.18) <.001
24-mo CESD −0.77 (−0.83 to −0.71) <.001
24-mo PSS −0.33 (−0.41 to −0.24) <.001
Multivariable model for 24-mo fatigue change (pretreatment minus 24 mo)
Intercept −32.44 (−34.24 to −30.64) <.001
Pretreatment fatigue 0.65 (0.62 to 0.69) <.001
24-mo CESD 0.68 (0.6 to 0.76) <.001
24-mo PSS 0.33 (0.24 to 0.41) <.001
6MWD per 10 m change from 3 to 24 mo 0.07 (0.05 to 0.10) <.001
Fatigue change from 0 to 3 mo 0.17 (0.15 to 0.19) <.001
CESD change from 0 to 24 mo −0.10 (−0.16 to −0.03) .003
Open in a new tab

Abbreviations: 6MWD, 6-minute walking distance; CESD, Center for Epidemiological Studies Depression Scale; LASSO, lease absolute shrinkage and selection operator; PSS, perceived stress scale.

a

Results from 2 independent multivariable linear regression models—1 for 24-month fatigue score and the other for 24-month change from pretreatment in fatigue. Included covariates were selected from parent models via LASSO regression and 10-fold cross validation.

Table 2 also presents results from a multivariable model for pre- to 24-month fatigue change over time. Pretreatment fatigue and worsening in fatigue from pretreatment to 3 months were associated with clinically significant change in fatigue at 24 months. Each 10-meter improvement in 6MWD from 3 to 24 months was associated with greater reduction in fatigue over time (β, 0.07; 95% CI, 0.05 to 0.10; P < .001). Larger reductions in depressive symptoms from pretreatment to 24 months were associated with less of an increase in fatigue over 24 months (β, −0.10; 95% CI, −0.16 to −0.03; P = .003). Change in PSS was not associated with the 24-month increase or decrease in fatigue scores in this multivariable model. A sensitivity analysis restricting the cohort to participants without prior chemotherapy, radiation, or hormonal therapy yielded materially unchanged results (eTable 7 in Supplement 1).

Mediation Analyses

From bivariate analyses (Figure 3; eTable 5 in Supplement 1) the only 2 significant candidate factors formally assessed as mediators were PSS and depressive symptoms. In a latent change score mediation model, treatment group assignment was associated with changes in both CESD and PSS, which in turn were associated with changes in participant-reported fatigue. Relative to the comparators, both chemotherapy groups experienced increased fatigue at 3 months, partially mediated by worsening CESD (indirect effect −1.16; 95% CI, −2.44 to 0.13 [14.5% of the total effect]; and −1.91; 95% CI, −3.33 to −0.49 [19.1%]). At 24 months, both groups reported significant improvements in fatigue from 3 to 24 months, substantially mediated by CESD improvement (indirect effect, 2.14; 95% CI, 0.77 to 3.51 [25.0%]; and 2.04; 95% CI, 0.67 to 3.41 [21.4%]). An additional 2% to 3% was mediated through joint CESD through PSS pathways.

Figure 3. Mediation Analysis Graph of Estimated Pathways Linking Treatment, Depression, Stress, and Fatigue Change Over Time.

Six-panel mediation path diagrams with coefficients for CESD, PSS, and fatigue. Six panels labeled A through F arranged in two rows of three. Each panel contains a left-to-right path diagram of four light blue boxes labeled Group, CESD, PSS, and Fatigue, connected by thin gray arrows Group to CESD, CESD to PSS, and PSS to Fatigue. Each panel also includes a large curved arrow across the top from Group toward Fatigue with a numeric estimate, 95 percent confidence interval, and P value printed above the arc. A second curved arrow along the bottom connects CESD toward Fatigue; in some panels this bottom arc is dashed. A short horizontal arrow from PSS to Fatigue has its own estimate and P value printed above it in each panel. Panel A title: Chemotherapy inclusive of doxorubicin, 0 to 3 mo. Top arc: minus 6.84, 95 percent C I minus 8.20 to minus 5.40; P equals .001. PSS to Fatigue: minus 0.14, 95 percent C I minus 0.31 to 0.05; P equals .16. Bottom arc dashed: minus 1.16, 95 percent C I minus 2.44 to 0.13; P equals .07. Text below: Total minus 8.02, 95 percent C I minus 10.81 to minus 5.23; P equals .001. Approximately 16.1 percent mediated. Panel B title: Chemotherapy without doxorubicin, 0 to 3 mo. Top arc: minus 8.04, 95 percent C I minus 9.40 to minus 6.70; P equals .001. PSS to Fatigue: minus 0.23, 95 percent C I minus 0.45 to minus 0.01; P equals .04. Bottom arc solid: minus 1.91, 95 percent C I minus 3.33 to minus 0.49; P equals .009. Text below: Total minus 9.97, 95 percent C I minus 12.80 to minus 7.14; P equals .001. Approximately 21.2 percent mediated. Panel C title: Aromatase inhibitors, 0 to 3 mo. Top arc: 1.94, 95 percent C I 0.70 to 3.20; P equals .002. PSS to Fatigue: minus 0.08, 95 percent C I minus 0.28 to 0.12; P equals .47. Bottom arc dashed: minus 0.68, 95 percent C I minus 2.21 to 0.85; P equals .46. Text below: Total 1.11, 95 percent C I minus 2.22 to 4.44; P equals .51. Panel D title: Chemotherapy inclusive of doxorubicin, 3 to 24 mo. Top arc: 6.13, 95 percent C I 4.70 to 7.60; P equals .001. PSS to Fatigue: 0.25, 95 percent C I 0.04 to 0.46; P equals .03. Bottom arc solid: 2.16, 95 percent C I 0.77 to 3.51; P equals .002. Text below: Total 8.57, 95 percent C I 5.84 to 11.30; P equals .001. Approximately 27.9 percent mediated. Panel E title: Chemotherapy without doxorubicin, 3 to 24 mo. Top arc: 7.31, 95 percent C I 5.90 to 8.70; P equals .001. PSS to Fatigue: 0.23, 95 percent C I 0.04 to 0.44; P equals .02. Bottom arc solid: 2.05, 95 percent C I 0.67 to 3.41; P equals .004. Text below: Total 9.51, 95 percent C I 6.75 to 12.28; P equals .001. Approximately 23.9 percent mediated. Panel F title: Aromatase inhibitors, 3 to 24 mo. Top arc: 1.23, 95 percent C I 0.00 to 2.50; P equals .05. PSS to Fatigue: 0.13, 95 percent C I

Open in a new tab

Estimated pathways linking treatment group, depressive symptoms (Center for Epidemiological Studies Depression scale [CESD]), perceived stress (Perceived Stress Scale [PSS]), and fatigue change. Top panels show 0 to 3 month associations; bottom panels show 3 to 24 month associations for chemotherapy with doxorubicin, chemotherapy without doxorubicin, and aromatase inhibitors with no chemotherapy. Solid arrows denote significant paths (P < .05); dashed arrows denote nonsignificant paths. Values are standardized estimates with 95% CIs and P values; total associations and proportion mediated (sum of CESD, fatigue and CESD, PSS, and fatigue indirect associations) are shown below each panel.

In an assessment of direct associations, changes in CESD and PSS were each independently associated with fatigue change at both time points, with worsening depression (3 months: β, −1.08; 95% CI, −1.33 to −0.83 and 24 months: β, −0.88; 95% CI, −1.12 to −0.64) and greater perceived psychosocial stress (3 months: β, −0.36; 95% CI, −0.68 to −0.05 and 24 months: β, −0.32; 95% CI, −0.62 to −0.01) significantly associated with greater fatigue burden (P < .05 for all paths). The significant perceived-stress pathway in the structural model contrasts with the LASSO change model (Table 2), where change in perceived stress was not retained. Because LASSO selects among collinear factors, the perceived-stress contribution is absorbed by the correlated change in depressive symptoms, whereas the latent change score model estimates it explicitly. Perceived stress is therefore best interpreted as a secondary contributor—smaller than depressive symptoms and not retained under penalized selection, although significant when modeled directly.

Discussion

The findings of this longitudinal prospective cohort study of 236 women receiving potentially cardiotoxic chemotherapy for stage I to III BC and 145 comparators without cancer recruited from community and academic health systems in the US showed first, that 2 years after receiving treatment for BC and after accounting for age, BC stage, and the presence of CV risk factors, 18.2% of women receiving potentially cardiotoxic chemotherapy experienced clinically relevant fatigue regardless of receipt of anthracycline or nonanthracycline based chemotherapy regimens. Also, 34.6% of patients receiving chemotherapy experienced clinically relevant increases in fatigue manifested by at least 5 points from baseline declines in the Functional Assessment of Chronic Illness Therapy Fatigue Scale.18 These percentages are consistent with other studies that have measured fatigue over a similar length of time.34,35 Second, that relative to women receiving nonanthracycline based chemotherapy or aromatase inhibitors no chemotherapy, women receiving anthracycline-based chemotherapy experienced baseline adjusted declines in 6MWD, LVEF, and LV circumferential strain, however none of these were associated with 24-month post–BC treatment fatigue. Additionally, that after accounting for the presence of pretreatment and posttreatment LVEF, diabetes, hypertension, tobacco use, hematocrit, age, body mass index, and receipt of potentially cardio-protective medications, depressive symptoms were the principal mediator of the association between BC treatment group on 24-month fatigue, with perceived stress contributing secondarily.

Participants in this study reflected similar demographics, risk factor profiles, anticancer and antihypertensive treatment regimens, and resting values of cardiac parameters to other populations reported across other studies involving women receiving treatment for BC (Table 1; eTable 2 in Supplement 1).36,37,38,39 Seventeen percent (and nearly 20% of those receiving doxorubicin) identified as Black, which is slightly greater relative to Black-identifying women enrolled in prior studies.40,41

As shown in Figure 2, 3 months into treatment for BC, fatigue worsened on average for women receiving anthracycline and nonanthracycline-based chemotherapeutic regimens, but not in women receiving AIs with no chemotherapy or those without BC. For women receiving chemotherapy at 3 months, there were declines in hematocrit, LV ejection fraction, LV circumferential strain, 6MWD, and systolic blood pressure, and increases in perceived stress and depressive symptoms. These changes occurred regardless of race, age, cancer stage, or receipt of cardioprotective medications, and many of these measures recovered at the 24-month post–BC initiation of treatment (eTable 3 in Supplement 1), and with the exception of perceived stress and depressive symptoms, none were associated with fatigue at 24 months (eTable 4 in Supplement 1).

Using various metrics, other studies have also shown a relationship between depressive symptoms or stress and fatigue.34,35,42,43,44,45,46,47,48 Previous research has also shown that stress is related to depression.49,50,51,52 However, none of these studies have included measures of cardiovascular function, fatigue, depression, and perceived stress altogether. Furthermore, to our knowledge, few studies have measured the extensive range of clinical parameters related to fatigue as in the present study. One exception is a cross-sectional study that examined 1787 participants with cancer and found that participants with higher levels of depressive symptoms, pain, and lower hemoglobin levels had significantly greater fatigue.43 In our analyses, depressive symptoms dominated, carrying most of the mediated contribution with associations roughly 3 times those for perceived stress. Interventions targeting depression may therefore reduce posttreatment fatigue, with stress a secondary target.

Surprisingly, there was a lack of an association of clinical CV risk factors (eg, hypertension, diabetes, smoking) or objective measures of cardiac function including LV ejection fraction or strain parameters with 24-month post initiation of BC treatment fatigue. Future studies may determine if immune, inflammatory, neuroendocrine, and neural processes including activation of the hypothalamic-pituitary-adrenal axis are associated with cancer-related fatigue.53,54 Since the American Society of Clinical Oncology has suggested physical activity, and psychosocial interventions, may associate with improvements in both depressive symptoms and perceived stress,55,56 our results support consideration of screening for depressive symptoms and the presence of stress. Depressive symptoms are common (ranging from 12% to 25%) among participants with BC,57,58,59,60 are associated with poorer quality of life and may contribute to increased mortality.61,62,63,64,65

Limitations

This study has some limitations. First, the study was conducted during the COVID-19 pandemic, and some participants had missed or delayed scheduled visits. A total of 61 of 381 participants dropped out before 24 months. However, 84% of the participants in the study had data across baseline, 3, and 24 months (Figure 1), and there was no significant difference in pretreatment fatigue between dropouts or those who remained at 24 months. Missing visit data were addressed with multiple imputation under a missing-at-random assumption; the imputation models included 34 demographic and risk-factor covariates, likely capturing factors of missingness. Complete-case and multiply imputed analyses yielded similar results. Second, although the study included mostly White participants, the percentage of Black participants exceeded that of many studies (66 [17.3% overall]) including those recruited from community hospitals where greater than 80% of women with BC receive treatment.5,6 Third, some potential mediators considered failed to show statistical significance. Importantly, the observed mediated proportions for the nonsignificant mediators were all less than 5%; even with a larger sample to detect them, such small associations would likely be clinically irrelevant. Fourth, as an observational cohort, our mediation estimates are interpretable as causal under assumptions of no unmeasured confounding and correct temporal ordering; we therefore report all findings as associations. Finally, although the noncancer comparator group does not represent a random sample, we believe this comparator sample controls for environmental and/or geographic exposures and to an extent may share some genetic characteristics of the patients with BC (if they are relatives). Of note, the primary results and conclusions of this study focus on the comparisons longitudinally within BC groups rather than across the BC vs comparator group. Therefore, the method of selection of the comparators likely has minimal impact on the overall conclusions.

Conclusions

In summary, the results of this study indicate that 34% of women receiving adjuvant and neoadjuvant chemotherapy for breast cancer experienced a clinically significant increase in fatigue 24 months after initiation of their treatment regardless of age, CV risk factors, or change in hematocrit or LV ejection fraction and/or circumferential strain. Importantly this BC treatment–related fatigue was partially mediated by symptoms of depression and the presence of stress irrespective of cardiovascular function. Thus, post–BC treatment of fatigue may require strategies targeting depression and stress independent of cardiovascular function.

Supplement 1.

eMethods. Statistical Analysis

eTable 1. Women With Breast Cancer Grouped by Treatment Type

eTable 2. Pretreatment Participant Characteristics by Cancer Treatment Group

eTable 3. Time Varying Measures by Cancer Treatment Group at Pretreatment, 3 Months, and 24 Months Follow Up

eTable 4. Association Between All Other Variables and Fatigue

eTable 5. Mediation Analyses

eTable 6. Development of Multivariable Model of Variables

eTable 7. Sensitivity Analyses of Multivariable Models Excluding Patients With Prior Chemotherapy, Chest Radiation, Hormonal Therapy, or Surgery

jamanetwopen-e2629862-s001.pdf (374.1KB, pdf)
Supplement 2.

Data Sharing Statement

jamanetwopen-e2629862-s002.pdf (15.9KB, pdf)

References

  • 1.Al Maqbali M. Cancer-related fatigue: an overview. Br J Nurs. 2021;30(4):S36-S43. doi: 10.12968/bjon.2021.30.4.S36 [DOI] [PubMed] [Google Scholar]
  • 2.Avis NE, Levine BJ, Klepin HD, et al. ; UPBEAT study team . The impact of non- and anthracycline-based chemotherapy on fatigue in breast cancer survivors: results from WF-97415. Support Care Cancer. 2024;32(8):528. doi: 10.1007/s00520-024-08717-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Abrahams HJG, Gielissen MFM, Schmits IC, Verhagen CAHHVM, Rovers MM, Knoop H. Risk factors, prevalence, and course of severe fatigue after breast cancer treatment: a meta-analysis involving 12 327 breast cancer survivors. Ann Oncol. 2016;27(6):965-974. doi: 10.1093/annonc/mdw099 [DOI] [PubMed] [Google Scholar]
  • 4.Nieboer P, Buijs C, Rodenhuis S, et al. Fatigue and relating factors in high-risk breast cancer patients treated with adjuvant standard or high-dose chemotherapy: a longitudinal study. J Clin Oncol. 2005;23(33):8296-8304. doi: 10.1200/JCO.2005.10.167 [DOI] [PubMed] [Google Scholar]
  • 5.Petrelli NJ. A community cancer center program: getting to the next level. J Am Coll Surg. 2010;210(3):261-270. doi: 10.1016/j.jamcollsurg.2009.11.015 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Pfister DG, Rubin DM, Elkin EB, et al. Risk adjusting survival outcomes in hospitals that treat patients with cancer without information on cancer stage. JAMA Oncol. 2015;1(9):1303-1310. doi: 10.1001/jamaoncol.2015.3151 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Amin MB, Greene FL, Edge SB, et al. The Eighth Edition AJCC Cancer Staging Manual: continuing to build a bridge from a population-based to a more “personalized” approach to cancer staging. CA Cancer J Clin. 2017;67(2):93-99. doi: 10.3322/caac.21388 [DOI] [PubMed] [Google Scholar]
  • 8.Copur MS, Ramaekers R, Gonen M, et al. ReCAP: impact of the National Cancer Institute Community Cancer Centers program on clinical trial and related activities at a community cancer center in rural Nebraska. JOP. 2016;12:67-68. doi: 10.1200/JOP.2015.005736 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Tucker TC, Charlton ME, Schroeder MC, et al. Improving the quality of cancer care in community hospitals. Ann Surg Oncol. 2021;28(2):632-638. doi: 10.1245/s10434-020-08867-y [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Jordan JH, D’Agostino RB Jr, Avis NE, et al. Fatigue, cardiovascular decline, and events after breast cancer treatment: rationale and design of UPBEAT study. JACC CardioOncol. 2020;2(1):114-118. doi: 10.1016/j.jaccao.2020.02.013 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Oken MM, Creech RH, Tormey DC, et al. Toxicity and response criteria of the Eastern Cooperative Oncology Group. Am J Clin Oncol. 1982;5(6):649-655. doi: 10.1097/00000421-198212000-00014 [DOI] [PubMed] [Google Scholar]
  • 12.Cleveland Clinic . Menopause. 2024. Accessed March 2026. https://my.clevelandclinic.org/health/diseases/21841-menopause
  • 13.Gold EB, Crawford SL, Avis NE, et al. Factors related to age at natural menopause: longitudinal analyses from SWAN. Am J Epidemiol. 2013;178(1):70-83. doi: 10.1093/aje/kws421 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.American Heart Association . Home blood pressure monitoring. 2025. Accessed November, 2025. https://www.heart.org/en/health-topics/high-blood-pressure/understanding-blood-pressure-readings/monitoring-your-blood-pressure-at-home
  • 15.Webster K, Cella D, Yost K. The Functional Assessment of Chronic Illness Therapy (FACIT) measurement system: properties, applications, and interpretation. Health Qual Life Outcomes. 2003;1:79. doi: 10.1186/1477-7525-1-79 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Yellen SB, Cella DF, Webster K, Blendowski C, Kaplan E. Measuring fatigue and other anemia-related symptoms with the Functional Assessment of Cancer Therapy (FACT) measurement system. J Pain Symptom Manage. 1997;13(2):63-74. doi: 10.1016/S0885-3924(96)00274-6 [DOI] [PubMed] [Google Scholar]
  • 17.Piper BF, Cella D. Cancer-related fatigue: definitions and clinical subtypes. J Natl Compr Canc Netw. 2010;8(8):958-966. doi: 10.6004/jnccn.2010.0070 [DOI] [PubMed] [Google Scholar]
  • 18.Cella D, Johansson P, Ueda Y, et al. Clinically important change for the FACIT-Fatigue scale in paroxysmal nocturnal hemoglobinuria: a derivation from international PNH registry patient data. J Patient Rep Outcomes. 2023;7(1):63. doi: 10.1186/s41687-023-00609-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Jordan JH, Sukpraphrute B, Meléndez GC, Jolly MP, D’Agostino RB Jr, Hundley WG. Early myocardial strain changes during potentially cardiotoxic chemotherapy may occur as a result of reductions in left ventricular end-diastolic volume: the need to interpret left ventricular strain with volumes. Circulation. 2017;135(25):2575-2577. doi: 10.1161/CIRCULATIONAHA.117.027930 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Bellenger NG, Davies LC, Francis JM, Coats AJ, Pennell DJ. Reduction in sample size for studies of remodeling in heart failure by the use of cardiovascular magnetic resonance. J Cardiovasc Magn Reson. 2000;2(4):271-278. doi: 10.3109/10976640009148691 [DOI] [PubMed] [Google Scholar]
  • 21.Forman DE, Fleg JL, Kitzman DW, et al. 6-min walk test provides prognostic utility comparable to cardiopulmonary exercise testing in ambulatory outpatients with systolic heart failure. J Am Coll Cardiol. 2012;60(25):2653-2661. doi: 10.1016/j.jacc.2012.08.1010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Andresen EM, Malmgren JA, Carter WB, Patrick DL. Screening for depression in well older adults: evaluation of a short form of the CES-D (Center for Epidemiologic Studies Depression Scale). Am J Prev Med. 1994;10(2):77-84. doi: 10.1016/S0749-3797(18)30622-6 [DOI] [PubMed] [Google Scholar]
  • 23.Cohen S, Kamarck T, Mermelstein R. A global measure of perceived stress. J Health Soc Behav. 1983;24(4):385-396. doi: 10.2307/2136404 [DOI] [PubMed] [Google Scholar]
  • 24.R Core Team . R: a language and environment for statistical computing. 2024. Accessed July 7, 2026. https://www.R-project.org/
  • 25.Rubin DB. Inference and missing data. Biometrika. 1976;63(3):581-592. doi: 10.1093/biomet/63.3.581 [DOI] [Google Scholar]
  • 26.van Buuren S, Groothuis-Oudshoorn K. mice: multivariate imputation by chained equations in R. Journal of Statistical Software. 2011;45(3):1-67. doi: 10.18637/jss.v045.i03 [DOI] [Google Scholar]
  • 27.O’Connell NS, Dai L, Jiang Y, et al. Methods for analysis of pre-post data in clinical research: a comparison of five common methods. J Biom Biostat. 2017;8(1):1-8. doi: 10.4172/2155-6180.1000334 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Friedman J, Hastie T, Tibshirani R. Regularization paths for generalized linear models via coordinate descent. J Stat Softw. 2010;33(1):1-22. doi: 10.18637/jss.v033.i01 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Tibshirani R. The LASSO method for variable selection in the Cox model. Stat Med. 1997;16(4):385-395. doi: 10.1002/(SICI)1097-0258(19970228)16:4<385::AID-SIM380>3.0.CO;2-3 [DOI] [PubMed] [Google Scholar]
  • 30.Rosseel Y. lavaan: an R package for structural equation modeling. J Stat Softw. 2012;48(2):1-36. doi: 10.18637/jss.v048.i02 [DOI] [Google Scholar]
  • 31.McArdle JJ. Latent variable modeling of differences and changes with longitudinal data. Annu Rev Psychol. 2009;60:577-605. doi: 10.1146/annurev.psych.60.110707.163612 [DOI] [PubMed] [Google Scholar]
  • 32.Selig JP, Preacher KJ. Mediation models for longitudinal data in developmental research. Res Hum Dev. 2009;6(2-3):144-164. doi: 10.1080/15427600902911247 [DOI] [Google Scholar]
  • 33.Grimm KJ, An Y, McArdle JJ, Zonderman AB, Resnick SM. Recent changes leading to subsequent changes: extensions of multivariate latent difference score models. Struct Equ Modeling. 2012;19(2):268-292. doi: 10.1080/10705511.2012.659627 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Rosas JC, Aguado-Barrera ME, Azria D, et al. ; REQUITE Consortium . (Pre)treatment risk factors for late fatigue and fatigue trajectories following radiotherapy for breast cancer. Int J Cancer. 2023;153(9):1579-1591. doi: 10.1002/ijc.34640 [DOI] [PubMed] [Google Scholar]
  • 35.Bower JE, Ganz PA, Irwin MR, et al. Do all patients with cancer experience fatigue? A longitudinal study of fatigue trajectories in women with breast cancer. Cancer. 2021;127(8):1334-1344. doi: 10.1002/cncr.33327 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.American College of Cardiology . New ACC/AHA high blood pressure guidelines lower definition of hypertension. 2017. Accessed October, 2023. https://www.acc.org/latest-in-cardiology/articles/2017/11/08/11/47/mon-5pm-bp-guideline-aha-2017
  • 37.Suerken CK, D’Agostino RB Jr, Jordan JH, et al. Simultaneous left ventricular volume and strain changes during chemotherapy associate with 2-year postchemotherapy measures of left ventricular ejection fraction. J Am Heart Assoc. 2020;9(2):e015400. doi: 10.1161/JAHA.119.015400 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Vaitiekus D, Muckiene G, Vaitiekiene A, et al. Impact of arterial hypertension on doxorubicin-based chemotherapy-induced subclinical cardiac damage in breast cancer patients. Cardiovasc Toxicol. 2020;20(3):321-327. doi: 10.1007/s12012-019-09556-3 [DOI] [PubMed] [Google Scholar]
  • 39.Sawaya H, Sebag IA, Plana JC, et al. Assessment of echocardiography and biomarkers for the extended prediction of cardiotoxicity in patients treated with anthracyclines, taxanes, and trastuzumab. Circ Cardiovasc Imaging. 2012;5(5):596-603. doi: 10.1161/CIRCIMAGING.112.973321 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Hundley WG, D’Agostino R Jr, Crotts T, et al. Statins and left ventricular ejection fraction following doxorubicin treatment. NEJM Evid. Published online August 18, 2022. doi: 10.1056/EVIDoa2200097 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Garg R, D’Agostino RB Jr, O’Connell N, et al. ; UPBEAT Study Team . Hypertension severity and declines in left ventricular ejection fraction among women receiving adjuvant chemotherapy for breast cancer (WF-97415 UPBEAT). Hypertension. 2024;81(6):1365-1373. doi: 10.1161/HYPERTENSIONAHA.123.21817 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Bower JE, Wiley J, Petersen L, Irwin MR, Cole SW, Ganz PA. Fatigue after breast cancer treatment: Biobehavioral predictors of fatigue trajectories. Health Psychol. 2018;37(11):1025-1034. doi: 10.1037/hea0000652 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Zeilinger EL, Zrnic-Novakovic I, Oppenauer C, et al. Prevalence and biopsychosocial indicators of fatigue in cancer patients. Cancer Med. 2024;13(11):e7293. doi: 10.1002/cam4.7293 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Blickle P, Schmidt ME, Steindorf K. Pretreatment fatigue in breast cancer patients: comparison with healthy controls and associations with biopsychosocial variables. Cancer Med. 2025;14(1):e70404. doi: 10.1002/cam4.70404 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Jakovljevic K, Kober KM, Block A, et al. Higher levels of stress are associated with a significant symptom burden in oncology outpatients receiving chemotherapy. J Pain Symptom Manage. 2021;61(1):24-31.e4. doi: 10.1016/j.jpainsymman.2020.07.019 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Ho RT, Kwan TT, Cheung IK, et al. Association of fatigue with perceived stress in chinese women with early stage breast cancer awaiting adjuvant radiotherapy. Stress Health. 2015;31(3):214-221. doi: 10.1002/smi.2548 [DOI] [PubMed] [Google Scholar]
  • 47.Wright F, Kober KM, Cooper BA, et al. Higher levels of stress and different coping strategies are associated with greater morning and evening fatigue severity in oncology patients receiving chemotherapy. Support Care Cancer. 2020;28(10):4697-4706. doi: 10.1007/s00520-020-05303-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Schaab M, Wijlens KAE, Bode C. Psychological coping factors associated with breast cancer-related fatigue: a systematic review of recent evidence for stages 0 to III. Clin Breast Cancer. 2023;23(7):e401-e411. doi: 10.1016/j.clbc.2023.06.005 [DOI] [PubMed] [Google Scholar]
  • 49.Langford DJ, Cooper B, Paul S, et al. Distinct stress profiles among oncology patients undergoing chemotherapy. J Pain Symptom Manage. 2020;59(3):646-657. doi: 10.1016/j.jpainsymman.2019.10.025 [DOI] [PubMed] [Google Scholar]
  • 50.Golden-Kreutz DM, Andersen BL. Depressive symptoms after breast cancer surgery: relationships with global, cancer-related, and life event stress. Psychooncology. 2004;13(3):211-220. doi: 10.1002/pon.736 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Andersen BL, DeRubeis RJ, Berman BS, et al. ; American Society of Clinical Oncology . Screening, assessment, and care of anxiety and depressive symptoms in adults with cancer: an American Society of Clinical Oncology guideline adaptation. J Clin Oncol. 2014;32(15):1605-1619. doi: 10.1200/JCO.2013.52.4611 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Seib C, Porter-Steele J, Ng SK, et al. Life stress and symptoms of anxiety and depression in women after cancer: The mediating effect of stress appraisal and coping. Psychooncology. 2018;27(7):1787-1794. doi: 10.1002/pon.4728 [DOI] [PubMed] [Google Scholar]
  • 53.Bower JE. The role of neuro-immune interactions in cancer-related fatigue: biobehavioral risk factors and mechanisms. Cancer. 2019;125(3):353-364. doi: 10.1002/cncr.31790 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54.Weber D, O’Brien K. Cancer and cancer-related fatigue and the interrelationships with depression, stress, and inflammation. J Evid Based Complementary Altern Med. 2017;22(3):502-512. doi: 10.1177/2156587216676122 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55.Bower JE, Bak K, Berger A, et al. ; American Society of Clinical Oncology . Screening, assessment, and management of fatigue in adult survivors of cancer: an American Society of Clinical oncology clinical practice guideline adaptation. J Clin Oncol. 2014;32(17):1840-1850. doi: 10.1200/JCO.2013.53.4495 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 56.Bellissimo MP, Canada JM, Jordan JH, et al. ; UPBEAT Study Group . Physical activity during breast cancer therapy associates with preserved exercise capacity and cardiac function (WF97415). JACC CardioOncol. 2023;5(5):641-652. doi: 10.1016/j.jaccao.2022.12.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 57.Avis NE, Levine BJ, Case LD, Naftalis EZ, Van Zee KJ. Trajectories of depressive symptoms following breast cancer diagnosis. Cancer Epidemiol Biomarkers Prev. 2015;24(11):1789-1795. doi: 10.1158/1055-9965.EPI-15-0327 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 58.Deshields T, Tibbs T, Fan MY, Taylor M. Differences in patterns of depression after treatment for breast cancer. Psychooncology. 2006;15(5):398-406. doi: 10.1002/pon.962 [DOI] [PubMed] [Google Scholar]
  • 59.Burgess C, Cornelius V, Love S, Graham J, Richards M, Ramirez A. Depression and anxiety in women with early breast cancer: five year observational cohort study. BMJ. 2005;330(7493):702. doi: 10.1136/bmj.38343.670868.D3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 60.Krebber AM, Buffart LM, Kleijn G, et al. Prevalence of depression in cancer patients: a meta-analysis of diagnostic interviews and self-report instruments. Psychooncology. 2014;23(2):121-130. doi: 10.1002/pon.3409 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 61.Bower JE. Behavioral symptoms in patients with breast cancer and survivors. J Clin Oncol. 2008;26(5):768-777. doi: 10.1200/JCO.2007.14.3248 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 62.Day R, Ganz PA, Costantino JP. Tamoxifen and depression: more evidence from the National Surgical Adjuvant Breast and Bowel Project’s Breast Cancer Prevention (P-1) Randomized Study. J Natl Cancer Inst. 2001;93(21):1615-1623. doi: 10.1093/jnci/93.21.1615 [DOI] [PubMed] [Google Scholar]
  • 63.DiMatteo MR, Lepper HS, Croghan TW. Depression is a risk factor for noncompliance with medical treatment: meta-analysis of the effects of anxiety and depression on patient adherence. Arch Intern Med. 2000;160(14):2101-2107. doi: 10.1001/archinte.160.14.2101 [DOI] [PubMed] [Google Scholar]
  • 64.Fann JR, Thomas-Rich AM, Katon WJ, et al. Major depression after breast cancer: a review of epidemiology and treatment. Gen Hosp Psychiatry. 2008;30(2):112-126. doi: 10.1016/j.genhosppsych.2007.10.008 [DOI] [PubMed] [Google Scholar]
  • 65.Mols F, Husson O, Roukema JA, van de Poll-Franse LV. Depressive symptoms are a risk factor for all-cause mortality: results from a prospective population-based study among 3,080 cancer survivors from the PROFILES registry. J Cancer Surviv. 2013;7(3):484-492. doi: 10.1007/s11764-013-0286-6 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplement 1.

eMethods. Statistical Analysis

eTable 1. Women With Breast Cancer Grouped by Treatment Type

eTable 2. Pretreatment Participant Characteristics by Cancer Treatment Group

eTable 3. Time Varying Measures by Cancer Treatment Group at Pretreatment, 3 Months, and 24 Months Follow Up

eTable 4. Association Between All Other Variables and Fatigue

eTable 5. Mediation Analyses

eTable 6. Development of Multivariable Model of Variables

eTable 7. Sensitivity Analyses of Multivariable Models Excluding Patients With Prior Chemotherapy, Chest Radiation, Hormonal Therapy, or Surgery

jamanetwopen-e2629862-s001.pdf (374.1KB, pdf)
Supplement 2.

Data Sharing Statement

jamanetwopen-e2629862-s002.pdf (15.9KB, pdf)

Articles from JAMA Network Open are provided here courtesy of American Medical Association

RESOURCES