Fax: (011) 49-6221-548142
A population-based study of the impact of specific symptoms on quality of life in women with breast cancer 1 year after diagnosis
Article first published online: 17 OCT 2006
Copyright © 2006 American Cancer Society
Volume 107, Issue 10, pages 2496–2503, 15 November 2006
How to Cite
Arndt, V., Stegmaier, C., Ziegler, H. and Brenner, H. (2006), A population-based study of the impact of specific symptoms on quality of life in women with breast cancer 1 year after diagnosis. Cancer, 107: 2496–2503. doi: 10.1002/cncr.22274
- Issue published online: 8 NOV 2006
- Article first published online: 17 OCT 2006
- Manuscript Accepted: 29 AUG 2006
- Manuscript Revised: 8 AUG 2006
- Manuscript Received: 30 MAY 2006
- German Cancer Foundation (Deutsche Krebshilfe). Grant Numbers: 70-1816, 70-2413
- breast neoplasms;
- health status;
- multivariate analysis;
- quality of life;
Whereas the role of specific symptoms, such as pain and fatigue, for quality of life (QOL) is unquestioned, their relative importance for long-lasting impairments in QOL in cancer patients has rarely been assessed quantitatively. The authors, therefore, aimed to identify symptoms most predictive of limitations to function and overall QOL in women with breast cancer after completion of primary therapy.
The European Organisation for Research and Treatment of Cancer questionnaire QLQ-C30 and the breast–cancer-specific module QLQ-BR23 were used to measure QOL in a population-based sample from Saarland (Germany) of 314 women with breast cancer 1 year after diagnosis. Symptoms most predictive for limitations to function and overall QOL were identified with a multiple linear regression analysis.
Fatigue emerged as the strongest predictor by far of QOL, explaining around 30% to 50% of variability within function scores and overall QOL. Other symptoms, including pain, nausea and/or vomiting, breast symptoms, systemic therapy side effects, and arm symptoms, explained on average <5% of variability of various QOL scales beyond fatigue and age. Sociodemographic and clinical factors had little impact on QOL.
Although QOL is a multidimensional concept, the analysis suggested that fatigue is the symptom that had, by far, the largest impact on limiting function and on overall QOL in breast cancer patients after their completion of primary therapy. Specific interventions to reduce the burden of fatigue may represent a particularly worthwhile effort to improve QOL in women with breast cancer. Cancer 2006. © 2006 AmericanCancer Society.
Health-related quality of life (QOL) is a multidimensional construct dependent on clinical, sociodemographic and psychological factors. Initially, QOL was assessed as an outcome measure in palliative care and as an additional endpoint in the evaluation of different treatment alternatives, but during the past decade, research has focused on disentangling the multidimensional mechanisms that may affect QOL in cancer patients after their completion of primary therapy.
Unquestionably, specific symptoms like pain, fatigue, sleeping disorders, and financial difficulties are of particular concern for women with breast cancer, and these symptoms may hamper their QOL even after the period of acute treatment.1–4 However, little is known about the importance of various breast cancer symptoms to impairment of function, and only a few studies have specifically addressed the impact of specific symptoms (such as fatigue, and arm or breast symptoms) on performance of life functions in cancer survivors.2, 3, 5–7 A better understanding of the relation between specific symptoms and limitations to function may provide a promising avenue that leads to development of specific interventions to improve QOL in women with breast cancer.
Therefore, in a large sample of breast cancer patients, we aimed to quantitatively assess the role that specific symptoms or health complaints play in specific impairments to function and in overall QOL.
MATERIALS AND METHODS
Our study is based on a population-based state-wide cohort of breast cancer patients from Saarland (a state in the southwest of Germany covering a population of 1 million inhabitants) that we initiated to study risk factors, diagnostic procedures, prognosis, and QOL in women with breast cancer.8, 9 In brief, 387 women with breast cancer diagnosed between October 1996 and February 1998 were recruited within days after diagnosis from all hospitals that offer in-patient cancer treatment in the study region and in adjacent counties. Study eligibility criteria for baseline recruitment included histologically confirmed invasive cancer, an age of 18 to 80 years, sufficient knowledge of the German language, and residency in the state of Saarland. After a comprehensive baseline interview (see below), patients were followed with respect to vital status and QOL 1 year after diagnosis. The protocols for this study were approved by local and regional ethics committees. Informed written consent was obtained from each patient.
Baseline data including year of birth, nationality, education, marital status, household size, as well as date of diagnosis and presence of comorbidity were derived from structured face-to-face interviews. The interviews were conducted by trained interviewers within a few days to weeks after diagnosis of breast cancer, in most cases during first hospitalization due to breast cancer. Other concurrent chronic diseases that we subsumed under a comorbidity index included myocardial infarction, congestive heart failure, peripheral vascular disease, cerebrovascular disease, dementia, chronic pulmonary disease, arthrosis, connective tissue disease, ulcerative disease, liver disease, diabetes, stroke, renal disease, other malign tumor, and acquired immune deficiency syndrome (AIDS).
Information regarding tumor stage at time of diagnosis and initial therapy was abstracted from hospital records. Information on further surgical and other therapeutic procedures, including regimen of adjuvant therapy, was obtained from hospital discharge reports and from the patient during follow-up. Information on disease progression or disease recurrence was provided by the Saarland Cancer Registry.
One year after their diagnosis of breast cancer, we mailed a QOL questionnaire (see below) to all study participants. Nonrespondents were mailed up to 2 reminders and were contacted by telephone if they did not respond after 3 mailings. If all these attempts did not result in any response, vital status of nonrespondents was obtained from the municipal registration offices.
QOL was assessed with the Quality of Life Questionnaire Core 30 Items (QLQ-C30) of the European Organisation for Research and Treatment of Cancer (EORTC)10 and the breast cancer-specific module QLQ-BR23 also developed by the EORTC.11 The QLQ-C30 is a validated, brief, self-reporting, cancer-specific measure of health-related QOL. It is composed of 5 multi-item function scales that evaluate physical, role, emotional, cognitive, and social function, and 1 global health status/quality-of-life scale. Three multi-item symptom scales measure fatigue, pain, and nausea and/or vomiting, and 6 single items assess symptoms such as dyspnea, insomnia, appetite loss, constipation, diarrhea, and financial difficulties. The QLQ-BR23 module incorporates multi-item scales to assess systemic therapy side effects, arm symptoms, breast symptoms, body image, sexual function, and single items to assess sexual enjoyment, upset by hair loss, and future perspective. The time frame for all scales in the questionnaire was the patient's past week, except for items related to sexual activity where a 4-week time frame was applied.
The scoring of the QLQ-C30 and QLQ-BR23 items was performed in accordance with the EORTC scoring manual.12 All scores were linearly transformed to a 0- to 100-points scale. In case of missing items, multi-item scores were calculated as the mean of nonmissing items if at least half of the items from the corresponding scale had been completed. In both instruments, high function scores represent better functioning and QOL, whereas high symptom scores indicate more severe symptoms. A cutoff value of 50 has been suggested by Koller and Lorenz to indicate clinically significant impairments.13
We calculated Pearson partial correlation coefficients to assess the association between symptom and function scores (including overall QOL) after adjustment for age and to discover which symptoms or health complaints may be responsible for impaired QOL in women with breast cancer. Multiple regression analysis with stepwise forward selection (P = .15) was applied to identify subsets of symptoms that were independently predictive for various function domains and overall QOL. All models were adjusted for age and were limited to the 5 most predictive symptoms. Partial R2-values were calculated to assess the independent proportion of explained variance within QOL by each symptom included in the model. For nominal variables, such as sociodemographic and treatment-related characteristics, analysis of variance (ANOVA) was employed to quantify the explained variance within the specific function domains of QOL.
Of 387 women, 373 survived the first year after tumor diagnosis (96%). Of these, 314 (84%) participated in the QOL survey and returned the questionnaire. Respondents and nonrespondents were similar with respect to age, education, and comorbidity status.8 However, tumor stage was more favorable among respondents (55% versus 48% localized disease), and respondents tended to live more often with a partner (66% versus 53%).
The mean age of study participants was 58.2 years (Table 1). More than 66% reported 1 or more other chronic health conditions. Involvement of regional lymph nodes was observed in 42% of all women, and distant metastasis in 5% of all women, including those women who experienced further disease progression during the first year after their diagnosis. Breast-conserving therapy (including secondary breast reconstruction) was performed in 57% of all women, and almost all women underwent axilla dissection (with 10 and more lymph nodes examined in more than 80% of all cases). Mode and extent of therapy varied across different age categories with breast-conserving therapy, axilla dissection, radiation, and chemotherapy less often performed in older women. Most patients undergoing chemotherapy were treated with cyclophosphamide/methotrexate/fluorouracil (39%), epirubicin/cyclophosphamide (34%), or a combination of both (25%).
|Total population (N = 314)||By age|
|<50 y (n = 79)||50-65 y (n = 146)||>65 y (n = 89)|
|Yes (all are adults)||71%||16%||88%||93%|
|Yes (at least 1 minor)||14%||51%||3%||0%|
|1 comorbidity condition||31%||34%||29%||30%|
|2+ comorbidity conditions||36%||13%||37%||54%|
Table 2 shows the distribution of the function and symptom scores measured at 1 year after diagnosis of breast cancer. The highest function scores were found for physical and cognitive functioning, whereas emotional functioning, sexual activity, and future perspective were rated much lower. For the latter 3 items, summary function scores lower than 50 were reported by 33%, 68%, and 60%, respectively, of all study participants. Fatigue, sleeping disorders, pain, and arm symptoms (including lymphedema, numbness, and tingling sensations) were the most frequently reported symptoms. Each of these 4 symptoms was scored ≥50 by at least 30% of all women.
|Scale||N||Mean||Standard deviation||Proportion of cases with|
|Function scores*||Scores < 50|
|Scales covered by QLQ-C30|
|Overall quality of life||312||65.3||22.0||16%|
|Scales covered by QLQ-BR23|
|Symptom scores‡||Scores ≥ 50|
|Scales covered by QLQ-C30|
|Scales covered by QLQ-BR23|
|Systemic therapy side effects||306||24.7||18.9||13%|
|Upset by hair loss||300||11.9||29.0||10%|
Table 3 shows that the symptom scores were highly correlated, even after adjustment for age. The QLQ-C30 mainly reflects physical and emotional problems but also includes financial difficulties. The partial correlation coefficient exceeded a value of 0.30 in more than 50% of all cells within the correlation matrix. In particular, fatigue, pain, systemic therapy side effects (eg, xerostoma, hot flushes, head ache, hair loss), and arm and breast symptoms were highly correlated.
|Questionnaire and item||FA||SL||PA||DY||CO||AP||DI||NV||FI||SYS||BR||ARM||HU|
|Sleeping disorders (SL)||—||0.39‡||0.35‡||0.30‡||0.30‡||0.09°||0.28‡||0.29‡||0.51‡||0.37‡||0.38‡||0.16†|
|Appetite loss (AP)||—||0.15†||0.55‡||0.27‡||0.36‡||0.18†||0.19†||0.12*|
|Financial difficulties (FI)||—||0.51‡||0.38‡||0.42‡||0.22‡|
|Systemic therapy side effects (SYS)||—||0.54‡||0.50‡||0.43‡|
|Breast symptoms (BR)||—||0.58‡||0.26‡|
|Arm symptoms (ARM)||—||0.20‡|
|Upset by hair loss (HU)||—|
With respect to function scores, women reporting high symptom scores tended to report lower function scores irrespective of age (Table 4). Fatigue, pain, systemic therapy side effects, and arm symptoms were identified as the symptoms most highly correlated with impairments to functions. Symptoms, such as sleeping problems, dyspnea, appetite loss, nausea and vomiting, breast symptoms (ie, tenderness, pain, skin irritation), and financial problems, were also highly correlated with most function scores but to a somewhat lesser degree than the 4 above-mentioned symptoms. The correlation was much weaker between 2 other symptoms measured by the QLQ-C30 and the QLQ-BR23, namely diarrhea and feeling upset about hair loss, and the various function scales.
|Symptom score||Overall QOL||Physical function||Emotional function||Social function||Role function||Cognitive function||Sexual function||Body image||Future perspective|
|Systemic therapy side effects||−0.50‡||−0.46‡||−0.60‡||−0.62‡||−0.40‡||−0.49‡||−0.22†||−0.42‡||−0.40‡|
|Upset by hair loss||−0.13*||−0.01°||−0.19†||−0.21†||−0.08°||−0.11°||−0.05°||−0.21†||−0.17†|
Results of the multivariate analysis are shown in Table 5. Fatigue emerged as the strongest predictor of impaired QOL. After adjustment for age, severity of fatigue explained approximately 30% to 50% of variability within function scales measured by the QLQ-C30. Other symptoms, including pain, nausea and/or vomiting, breast symptoms, systemic therapy side effects, and arm symptoms, explained very little of the remaining variance of various QOL scores of the QLQ-C30 beyond fatigue and age. Overall, age and the 5 most predictive symptoms accounted for 40% to 60% of variability within function scores of the QLQ- C30 as opposed to 25% to 33% of variability within breast cancer-specific function scores of the QLQ-BR23.
|QOL scale||Step 0*||Step 1||Step 2||Step 3||Step 4||Step 5||Total R2|
|Overall QOL||Age 0.013||Fatigue 0.392||Pain 0.064||Financial diff. 0.021||Appetite loss 0.015||Arm symptoms 0.004||0.509|
|Physical function||Age 0.063||Fatigue 0.448||Nausea/Vomiting 0.037||Pain 0.013||Systemic therapy side effects 0.013||Appetite loss 0.012||0.586|
|Emotional function||Age 0.027||Fatigue 0.469||Breast symptoms 0.055||Insomnia 0.015||Pain 0.005||Appetite loss 0.004||0.575|
|Social function||Age 0.015||Fatigue 0.437||Systemic therapy side effects 0.046||Appetite loss 0.030||Pain 0.008||Financial diff. 0.005||0.541|
|Role function||Age 0.011||Fatigue 0.333||Arm symptoms 0.042||Appetite loss 0.017||Constipation 0.001||Dyspnea 0.001||0.421|
|Cognitive function||Age 0.017||Fatigue 0.333||Nausea/Vomiting 0.050||Breast symptoms 0.023||Dyspnea 0.011||Appetite loss 0.006||0.439|
|Sexual function||Age 0.170||Pain 0.065||Financial diff. 0.013||Dyspnea 0.009||Constipation 0.007||Diarrhea 0.004||0.266|
|Body image||Age 0.140||Systemic therapy side effects 0.173||Fatigue 0.026||Financial diff. 0.020||Insomnia 0.001||Appetite loss 0.005||0.248|
|Future perspective||Age 0.053||Fatigue 0.183||Breast symptoms 0.064||Financial diff. 0.015||Arm symptoms 0.001||Diarrhea 0.003||0.327|
Additional analyses examining the impact of sociodemographic and clinical factors on QOL revealed that only a few of these factors were related to QOL (Table 6). Among these factors, age, tumor stage, and comorbidity emerged as the most prominent determinants of QOL; however, except for sexual activity, they explained only ≤10% of the observed variance within different aspects of QOL. Influence of other sociodemographic (such as family status, employment status, and education) and treatment-related factors was even smaller and was limited to specific dimensions.
|Factor||Overall QOL||Physical function||Emotional function||Social function||Role function||Cognitive function||Sexual activity||Body image||Future perspective|
|Type of surgery§||<0.01°||<0.01°||<0.01°||<0.01°||0.01°||<0.01°||<0.01°||0.10‡||<0.01°|
Diagnosis and treatment of breast cancer can affect psychological, sexual, and physical functioning to a great extent.14 To our knowledge, this is the first study to provide a comprehensive quantitative analysis of the contribution of various symptoms to overall QOL and its specific dimensions. Among the various symptoms examined, fatigue emerged as the strongest predictor, by far, of QOL and explained approximately 30% to 50% of variability within function scores and overall QOL. In contrast, sociodemographic and clinical factors had little impact on QOL. As fatigue is quite common in women with breast cancer, efforts to reduce fatigue may have a large potential to improve overall QOL in women with breast cancer.
Fatigue, sleeping disorders, pain, and arm symptoms were the most frequently reported symptoms. This is consistent with a recent study that found fatigue, pain, and insomnia as the most prevalent and severe symptoms in women with breast cancer after their completion of surgical treatment.15 Although most acute treatment-related symptoms (such as nausea, vomiting, hair loss) decline during the first year after diagnosis, some intermediate and late effects, such as pain, fatigue, lymphedema, weight gain, and menopausal symptoms, may persist and are likely to have a detrimental effect on QOL and the function status of women with breast cancer.16, 17
In general, women reporting high symptom scores tended to report lower function scores and lower overall QOL. Among all symptoms considered, fatigue emerged as the strongest predictor of impaired QOL and explained approximately 30% to 50% of variability within function scales measured by the QLQ-C30. Although fatigue is considered to be 1 of the most common symptoms in cancer patients,18 its substantial impact on QOL of oncology patients is still under-recognized.19 This is even more surprising because fatigue has been shown to have a negative impact on overall QOL among breast cancer survivors,1, 20–22 and it has been associated with a number of psychosocial, physical, and medical conditions, such as depression, anxiety, pain, nausea, weakness, and appetite loss.21, 23–26 However, little is known about the etiology of cancer-related fatigue.27 Although fatigue is more common among women during adjuvant chemotherapy, conflicting study results have not yet shown whether these patients are at higher risk of developing persistent fatigue.21, 22, 28 Similarly, no clear relation appears to exist between sociodemographic factors and the occurrence of fatigue,21 thus limiting the identification of high-risk groups of patients.
Given that a profound communication gap seems to exist between patient and physician regarding fatigue and given the observation that fatigue is associated with significant physical, emotional, psychological, and social consequences, it has been suggested that patients may benefit from physician-initiated discussion of causes of and treatments for fatigue and that physicians may benefit from education regarding available treatment modalities.29 Mild physical exercise, sleep hygiene, and attention-restoring activities have been suggested to reduce severity of fatigue in cancer patients.30–32 In addition, prescription of psychostimulants and erythropoietin or transfusion of red blood cells have been proposed, but evidence on their effectiveness to improve fatigue and QOL is limited.30–33
In addition to fatigue, chronic pain after mastectomy or lumpectomy with axillary node dissection, which is reported by about 20% to 30% of all women, has been found to interfere with both occupational and domestic activities.34 Similarly, more than one-third of all patients with conventional axillary dissection after breast cancer surgery suffer persistent arm problems, which may substantially affect QOL.2, 3, 35, 36
In light of the strong interrelation between various symptoms, including fatigue, sleeping disorders, and pain, it is hard to judge whether interventions specifically aimed to reduce fatigue will also substantially improve QOL throughout a broad range of various domains or whether a combined approach is more effective.25 However, it has been shown that patients whose fatigue improved reported substantially greater improvements in energy, ability to perform usual activities, and overall health.37
Studies investigating sociodemographic and clinical predictors of QOL among breast cancer survivors have identified several candidates for poorer QOL, such as younger age, lower level of education, greater severity of disease, axillary dissection, adjuvant therapy, and living alone.5, 7, 14, 38–43 Age, tumor stage, and comorbidity emerged as the most prominent determinants of QOL in our study; however, they explained very little of the observed variance within different aspects of QOL. Influence of other sociodemographic and treatment-related factors was even smaller and was limited to specific dimensions. As the QLQ-C30 has been developed primarily to monitor QOL in the context of clinical trials, it may be less sensible to depict subtle persistent treatment-related effects several months after completion of therapy. In addition, other factors, such as coping behavior and social support, may also be important determinants of QOL,44 and these factors may be interfering with sociodemographic and clinical factors.
We were able to achieve a high participation rate by multiple approaches. The use of an internationally validated instrument, the state-wide recruitment of a large, unselected sample of women with breast cancer, and a high proportion of complete of data are among the strengths of the current study that increase the representativeness of our results.
As the number of cancer survivors is expected to rise in upcoming years, the management of late and long-term sequelae of cancer and cancer treatment is becoming more and more important. Our analysis indicates that fatigue has a substantial impact on QOL in women with breast cancer at 1 year after diagnosis. Its impact on QOL by far exceeds the impact of any other disease-related symptom. These findings suggest that efforts to reduce specifically the burden of fatigue may be very promising approaches to enhance QOL in breast cancer survivors.
This study was supported by 2 grants from the German Cancer Foundation (Deutsche Krebshilfe), Project No. 70-1816, 70-2413
- 12The EORTC QLQ-C30 Scoring Manual. 2nd ed. Brussels: European Organization for Research and Treatment of Cancer; 1999., , , , , for the EORTC Quality of Life Study Group.
- 30National Cancer Institute. Fatigue. PDQ Cancer Information Summaries. Available online at: URL: http://cancer.gov/cancertopics/pdq/supportivecare/fatigue [accessed August 3, 2006].