Abstract
Background. An inverse association either between depression or impaired functional status and quality of life (QoL) has been reported for cancer patients, but the independent effect of depression or depressive symptoms and of functional impairment on QoL is unclear. Patients and methods. We investigated the prevalence of depression or depressive symptoms with the Beck Depression Inventory (BDI), the functional impairment with the ECOG-Performance-Status (ECOG-PS) and the QoL with the EORTC-QLQ-C30 questionnaire in a sample of 175 hospitalised cancer patients prior to the start of chemotherapy. Results. Sixteen of 175 patients (9.1%) screened positive for major depression, 29 (16.6%) had mild to moderate depressive symptoms. In 11 of 15 scales of the EORTC-QLQ-C30 questionnaire depression or depressive symptoms were significantly associated with worse QoL in univariate analysis and in 12 of 15 scales poor ECOG-PS was significantly associated with worse QoL. In multivariate analysis including ECOG-PS and BDI, the effect of depression and/or depressive symptoms on QoL was persistent in seven scales: global QoL, physical- and role functioning, fatigue, nausea & vomiting, pain, and constipation, that of ECOG-PS in five scales: global QoL, emotional functioning, nausea & vomiting, pain, and appetite loss. Conclusions. Signs of major depression or depressive symptoms and impaired functional status contribute independently to poorer QoL in cancer patients prior to chemotherapy.
Cancer is one of the world's most common diseases Citation[1]. Due to demographic changes, the number of cancer cases will increase within the next few decades Citation[2]. Often cancer is a severe and life threatening disease. Treatment may be harmful and results may be uncertain. Therefore functional abilities, presence of depression and depressive symptoms and quality of life (QoL) are important topics within the care for cancer patients.
The measurement of functional status/performance status (PS) has been established in oncology to rate patients overall fitness and the ability to care for themselves. Functional impairment or poor PS are of adverse prognostic importance for overall survival, especially with regard to early death and treatment-related toxicity Citation[3]. A number of studies have reported a close interaction between PS and QoL Citation[4], Citation[5].
Depression is a major symptom cancer patients suffer from. Masie provides a systematic review on the prevalence of major depression and depressive symptoms in cancer patients. She reports a prevalence of major depression of 0–38%, and of 0–58% for depressive symptoms in cancer patients Citation[6]. Risk factors of depression or depressive symptoms in cancer patients include functional disability, inadequate social support, uncontrolled pain, poor physical condition, advanced illness, previous history of depression, and the loss of a spouse or of family members Citation[7]. Presence of depression in cancer patients has been linked to a reduced chance of survival in a variety of tumours, e.g. breast cancer Citation[8] and ovarian cancer Citation[9]. A number of studies reported a close interaction between depression and QoL in cancer patients Citation[10], Citation[11].
QoL is a major area of concern in the treatment of patients with cancer Citation[12], especially in those treated within a non-curative approach. There are different questionnaires available to assess patients QoL. One of the most widely used in cancer patients is the EORTC-QLQ-C30 Citation[13]. A number of studies reported on the interaction of functional status and QoL (see above). A current PubMed research revealed only six citations for the quotes “EORTC-QLQ-C30” and “Beck Depression Inventory” (8th August 2006). These articles reported an inverse association either between depression or depressive symptoms and QoL.
Thus the association between functional impairment and poor QoL and between depression or depressive symptoms and poor QoL is established, but the independent effect of depression or depressive symptoms and of functional impairment on QoL is unclear. The association between global QoL, functional status and depression has been described by our group earlier Citation[14], but a detailed report on different scales of the EORTC-QLQ-C30 is missing so far.
Against this background, we therefore analysed data of depression and depressive symptoms, functional impairment and QoL measurement collected within a prospective trial. We asked to what extent depression or depressive symptoms contribute to different scales of QoL in addition to functional impairment, and vice versa.
Methods
The study was conducted in the Department of Haematology and Oncology at the University Hospital Jena, Germany. It was approved by the local ethical committee. Written informed consent was obtained by the treating physician after patients had been informed about the diagnosis of cancer and received the recommendation to undergo chemotherapy.
Patients
Of 427 patients aged 18 + years, newly admitted to the hospital in order to undergo chemotherapy for cancer and included in a clinical trial on decision making in cancer, 213 (49.9%) agreed to have an additional psychiatric evaluation. The first patient was included 06/12/1999, the last 29/08/2005. For 175–41.0% of all patients and 81.7% of those with psychiatric evaluation – all data on ECOG-PS, BDI and EORTC-QLQ-C30, were available. The following data were extracted from the patients’ records: (1) sex, (2) age, (3) tumour category – classified as solid or haematological – (4) diagnosis and (5) treatment approach – curative vs. non-curative –. Treatment approach was judged by the responsible physician based on tumour characteristics, (diagnosis, histology, stage) and patients’ characteristics (age, functional status, comorbidity).
Functional status
The most widely used scale to measure PS are the Karnofsky-PS (KPS) Citation[15] and the Eastern Co-operative Oncology Group (ECOG)-PS. Bucheri et al. favoured the ECOG-PS as prognostic tool compared to the KPS Citation[3], which therefore was selected for the trial. ECOG-PS consists of a 5 point scale: 0 = asymptomatic, 1 = symptomatic but completely ambulant, 2 = symptomatic, <50% in bed during the day, 3 = symptomatic, >50% in bed, but not bedbound, 4 = bedbound. Patients were grouped into those with a good ECOG-PS (0–1) and into those with a poor ECOG-PS (2–4). The ECOG-PS was rated by the trial physician.
Depression
Widely used instruments for screening are the Beck Depression Inventory (BDI) Citation[16], the Hospital Anxiety and Depression Scale (HADS) Citation[17], and the Structured Clinical Interview (SCID) according to the Statistical Manual of Mental Disorders (DSM-IV) Citation[18], Citation[19]. The BDI was originally developed for the use in psychiatric patients. Later on, it was used in other disease contexts Citation[20] and in the general population Citation[21]. Due to its world-wide acceptance we selected the BDI for screening of depression within this trial. The BDI contains 21 items; each can be answered on a four point Likert-scale between 0 to 3 according to the intensity of the item within the last week. The total score is obtained from the sum scores of the 21 items. A sum score of 0–10 is a normal result, 11–17 mild to moderate depressive symptoms and 18+ a clinical relevant depression. It is translated in a variety of languages. A detailed English manual is provided by Beck, a German version by Hautzinger et al. Citation[22], Citation[23]. The questionnaire was filled in interview-based by a trained rater. Results were checked by a professional.
Questionnaire for assessment of quality of life
The EORTC-QLQ-C30 is the most widely used instrument to measure QoL in cancer patients and was therefore selected for the use in this trial. We used the version 2 Citation[13]. Calculation of the scores was in accordance with the EORTC-QLQ-C30 scoring manual. For this trial the questionnaire was self administered. Only in patients with difficulties in reading the questionnaire was filled in interview-based by the trial physician.
Statistics
Data management and data analysis were performed with the statistical packages SPSS® 12 and SAS® 8.02. Checks for completeness, rightness, consistency, and plausibility were performed. Correlation was measured by Spearman-Correlation-Coefficient. P-values of differences between mean scores in BDI were calculated according to Mann-Whitney U-Test (n = 2 independent groups) and Kruskal-Wallis-Test (n ≥ 3 independent groups). The result of a statistic test with p-value <0.05 is called significant and with p-value <0.10 a trend. For multivariate analysis ANOVA was performed. Dependent variables in the model were the 15 scales of the EORTC-QLQ-C30, independent variables ECOG-PS and BDI. Due to the problem of multiple testing the results of the study were interpreted in an explorative manner.
Results
In total 175 patients were included in the analysis. Patients’ characteristics are presented in . More male than female patients were included, corresponding to the higher incidence rate of cancer in male than in female patients. Median age was 59.3 years (range 18–88 years). Patients aged 60+ were more often treated with a non-curative treatment approach than younger patients (75.6% vs. 45.5%; p < 0.001). 130 (74.4%) patients had no depression or depressive symptoms, 29 (16.6%) suffered from mild to moderate depressive symptoms, and 16 (9.1%) had signs of major depression in BDI. About two third of the patients had a good ECOG-PS (=0–1): ECOG-PS was 0 in 26%, 1 in 43%, 2 in 48%, and 3 in 4%.
Table I. Patients characteristics and BDI1 Score (mean, standard deviation). (Significant results are presented in bold).
Female sex, a solid tumour and a poor PS were significantly associated with higher BDI scores. Age and treatment approach did not show any significant influence.
reports the detailed single item scale response to the EORTC-QLQ-C30 for all patients, those with normal BDI, mild to moderate depressive symptoms, and signs of major depression. In addition it reports the association between single scales of the EORTC-QLQ-C30 in patients with good functional status (ECOG-PS = 0–1) and in those with poor functional status (ECOG-PS = 2–4). As comparison the normative values of a German standard population reported by Schwarz and Hinz Citation[24] are provided.
Table II. EORTC-QLQ-C30 scales mean score (standard deviation) for all patients, according to BDI categories (normal, mild to moderate depressive symptoms, signs of major depression), according to ECOG-Performance-Status (PS) (good PS vs. poor PS), and in comparison to a German normative sample of the EORTC-QLQ-C30 [24]. (Significant results are presented in bold).
A weak but significant correlation could be observed between BDI sum score and ECOG score (r = 0.24). A significant correlation between BDI sum score and QoL was found in 10 of 15 QoL scales: global QoL (r = − 0.43), physical functioning (r = -0.29), role functioning (r = − 0.72), emotional functioning (r = − 0.38), cognitive functioning (r = − 0.32), fatigue (r = 0.32), nausea and vomiting (r = 0.21), pain, insomnia (r = 0.35), and appetite loss (r = 0.36). Regarding the correlation between ECOG-PS and QoL scales, we found a significant correlation within the following 10 scales: global QoL (r = − 0.39), physical functioning (r = − 0.50), role functioning (r = − 0-37). cognitive functioning (r = − 0.25), fatigue (r = 0.37), nausea and vomiting (r = 0.25), pain (r = 0.33), dyspnoe (r = 0.24), appetite loss (r = 0.29), and constipation (r = 0.31).
To study the independent effect of both BDI and ECOG-PS, an analysis of variance (ANOVA) of the 15 scales of EORTC-QLQ-C30 and BDI and ECOG-Performance-Status (both continuous) was performed. demonstrates p-values and multiple correlation coefficients for each scale.
Table III. Results of analysis of variance (ANOVA) for to the different scales of quality of life according to ECOG-Performance-Status (ECOG-PS), Beck Depression Inventory (BDI) and interaction of BDI and ECOG-Performance-Status (if a significant variance on the different scales of QoL through ECOG-PS or BDI or its interaction exists, results are presented in bold).
In five scales the influence of ECOG-PS on QoL persisted, when adjusting for BDI: global QoL, emotional functioning, nausea & vomiting, pain, and appetite loss. In seven scales, the influence of BDI persisted, when adjusting for ECOG-PS: global QoL, physical- and role functioning, fatigue, nausea & vomiting, pain, and constipation. A significant interaction of BDI and ECOG-PS was observed for global QoL only.
Discussion
Cancer patients in general, and those treated with a non-curative approach in particular, regard the maintenance or regain of QoL as an important goal in cancer treatment Citation[25]. Consequently, it is interesting to know, which factors contribute to QoL in these patients. We provide data on QoL collected in cancer patients after diagnosis but prior to the start of chemotherapy. As expected the mean values of QoL were well below that of a German normative sample, consisting of 2 028 people selected by random-route-technique, age 19–82 years, mean age 49.4 years Citation[24]. And they are quite similar to data of other cancer patients, e.g. those described by Aaronson et al. in their initial report of the EORTC-QLQ-C30 Citation[13]. Results of EORTC-QLQ-C30 measurement in non-cancer patients are not reported so far.
The reported values of BDI are well above the mean values 6.45 (SD 5.2, range 0–19) reported for a German healthy control group, consisting of 86 persons, without psychosomatic or psychiatric disorders and although otherwise healthy, median age 55 years Citation[22]. The rates of depression (9.1%) and depressive symptoms (16.6%) found in our population are in line, with that described by other authors in cancer patients and summarised by Masie Citation[6].
Our data report a good PS in 68% of the patients. This is slightly lower than that in other patients series, e.g. in a cohort of cancer patients aged 65+ were 74% of patients had a good ECOG-PS Citation[26] or in that by Aaronson et al. were 71% had a good ECOG-PS Citation[13]. This might be explained by the fact, that we included only in-patients. Data on ECOG-PS in normative samples do not exist as comparison.
All in all the data on QoL, BDI, and ECOG-PS reported in our patients are comparable to other cancer patient series.
Poor PS was significantly associated with decreased QoL in 12 of 15 scales of the EORTC-QLQ in our data. This association between PS and QoL was reported by a number of authors for cancer and non-cancer patients as well Citation[4], Citation[5], Citation[13], Citation[27].
We observed a significant association of the presence of depression or depressive symptoms with decreased QoL in 11 of 15 scales of the EORTC-QLQ. The association of depression and QoL has been described for a variety of chronic diseases Citation[20], Citation[28], Citation[29], but is less well studied in cancer patients Citation[30]. Charlson et al. reported a correlation of BDI with most scales of the EORTC-QLQ-C30, however they included 17 patients only Citation[31]. Birkhauq et al. measured QoL with the EORTC-QLQ-C30 and the BDI in 200 laryngectomized patients. High BDI scores were associated with reduced QoL in most scales Citation[32]. We therefore provide one of the few reports on an association of depression and depressive symptoms and poor QoL in cancer patients.
Thus the question remained, whether functional impairment and presence of depression independently contribute to poor QoL, for which no data are reported previously. We could observe an independent association of BDI on seven and of ECOG-PS on five different scales of the EORTC-QLQ-C30. To some extend BDI, ECOG-PS and QoL overlap with each other and contribute to the overall subjective health condition of the patient. However an at least in partly independent contribution of impaired functional status and presence of depression or depressive symptoms on decreased QoL remains, as BDI and ECOG-PS show only a weak correlation (r = 0.24) and as results of analysis of variance demonstrate, that BDI and ECOG-PS measurement are not congruent to each other.
The data however have some limitations. The validation of the EORTC included the ECOG-PS, but not the BDI, the validation of the BDI did not include the ECOG-PS, thus we can not exclude some item bias within the results. As the group of cancer patients is heterogeneous, including patients with different diagnosis, different stages, different treatment approaches, and different time since diagnosis, the results might not suit for all of them. The distributions of the different diagnoses do not well reflect their frequency in the general population. With 51.3% of the patients suffering from haematological type of cancer the relative frequency of these patients is well above that in the general population with cancer. However the mean BDI scores were significantly higher in patients with solid tumour than with a malignant haematological disorder in our patients. We included only hospitalised patients, who might be more severely affected by the disease, than patients treated in an out-patients setting, and we included a high number of patients with non-curative treatment approach within our group of patients, but the mean BDI scores did not significantly differ between them. In addition we can not exclude some selection bias, as only 50% of the initial population agreed to a further psychiatric evaluation and in only 41% of all patients a full set of data was available. Thus the data might need validation in homogeneous patients’ samples.
Having these limitations in mind, we still provide the first report on an at least in partly independent association of depression or depressive symptoms and poor PS with different scales of the EORTC-QLQ in cancer patients. The results can have implications for cancer care and future research, as improvement of QoL is a major goal within the care for cancer patients and as the therapeutic approach towards depression and functional impairment is different. The data shall stimulate future research addressing the topic, how changes in BDI, e.g. by psychotherapeutic intervention and/or drug therapy, and changes in functional status, e.g. by effective cancer treatment, result in an improvement of QoL.
Conflict of interest
U.W. served as medical expert for Roche, Ortho-Biotec and Amgen, and received financial support for research from Ortho-Biotec, L.P. served as medical expert for Roche and Ortho-Biotec. A.K., B.R., H.S, K.H, and I.M. have no financial or personal relationship with other people or organisations that could inappropriately influence the presented work.
Acknowledgements
The study was supported by German Cancer Aid (Grant No. 70-2445-Hö-3). U.W. is currently research fellow of the Robert Bosch Foundation Stuttgart, Germany.
References
- Parkin DM, Bray F, Ferlay J, Pisani P. Global cancer statistics, 2002. CA Cancer J Clin 2002; 55: 74–108
- Edwards BK, Howe HL, Ries LA, Thun MJ, Rosenberg HM, Yancik R, et al. Annual report to the nation on the status of cancer, 1973–1999, featuring implications of age and aging on U.S. cancer burden. Cancer 1973; 94: 2766–92
- Buccheri G, Ferrigno D, Tamburini M. Karnofsky and ECOG performance status scoring in lung cancer: A prospective, longitudinal study of 536 patients from a single institution. Eur J Cancer 1996; 32A: 1135–41
- Schaafsma J, Osoba D. The Karnofsky Performance Status Scale re-examined: A cross-validation with the EORTC-C30. Qual Life Res 1994; 3: 413–24
- Jordhoy MS, Fayers P, Loge JH, Saltnes T, Ahlner-Elmqvist M, Kaasa S. Quality of life in advanced cancer patients: The impact of sociodemographic and medical characteristics. Br J Cancer 2001; 85: 1478–85
- Massie MJ. Prevalence of depression in patients with cancer. J Natl Cancer Inst Monogr 2004; 32: 57–71
- Breitbart W, Rosenfeld B, Pessin H, Kaim M, Funesti-Esch J, Galietta M, et al. Depression, hopelessness, and desire for hastened death in terminally ill patients with cancer. JAMA 2000; 284: 2907–11
- Watson M, Haviland JS, Greer S, Davidson J, Bliss JM. Influence of psychological response on survival in breast cancer: A population-based cohort study. Lancet 1999; 354(9187)1331–6
- Freyer G, Geay JF, Touzet S, Provencal J, Weber B, Jacquin JP, et al. Comprehensive geriatric assessment predicts tolerance to chemotherapy and survival in elderly patients with advanced ovarian carcinoma: A GINECO study. Ann Oncol 2005; 16: 1795–800
- Smith EM, Gomm SA, Dickens CM. Assessing the independent contribution to quality of life from anxiety and depression in patients with advanced cancer. Palliat Med 2003; 17: 509–13
- Grassi L, Indelli M, Marzola M, Maestri A, Santini A, Piva E, et al. Depressive symptoms and quality of life in home-care-assisted cancer patients. J Pain Symptom Manage 1996; 12: 300–7
- Bottomley A, Vanvoorden V, Flechtner H, Therasse P. The challenges and achievements involved in implementing Quality of Life research in cancer clinical trials. Eur J Cancer 2003; 39: 275–85
- Aaronson NK, Ahmedzai S, Bergman B, Bullinger M, Cull A, Duez NJ, et al. The European Organization for Research and Treatment of Cancer QLQ-C30: A quality-of-life instrument for use in international clinical trials in oncology. J Natl Cancer Inst 1993; 85: 365–76
- Pinquart M, Koch A, Eberhardt B, Brix C, Wedding U, Rohrig B. Associations of functional status and depressive symptoms with health-related Quality of Life in cancer patients. Qual Life Res 2006.
- Karnofsky DA, Adelmann WH, Craver FL. The use of nitrogen mustard in the palliative treatment of carcinoma. Cancer 1948; 1: 634–56
- Beck AT, Ward CH, Mendelson M, Mock J, Erbaugh J. An inventory for measuring depression. Arch Gen Psychiatry 1961; 4: 561–71
- Zigmond AS, Snaith RP. The hospital anxiety and depression scale. Acta Psychiatr Scand 1983; 67: 361–70
- Wittchen H-U, Wunderlich U, Gruschwitz S, Zaudig M. SKID-I Strukturiertes Klinisches Interview für DSM-IV, Achse I. Psychische Störungen. Hogrefe, Göttingen 1997
- First M, Gibbon M, Spitzer R, Williams J. Users guide for structured clinical interview for DSM-IV Axis 1 disorders-Research version. American Osychiatric Press, Inc., Washington, DS 1996
- Wilson B, Spittal J, Heidenheim P, Herman M, Leonard M, Johnston A, et al. Screening for depression in chronic hemodialysis patients: Comparison of the Beck Depression Inventory, primary nurse, and nephrology team. Hemodial Int 2006; 10: 35–41
- Oliver JM, Simmons ME. Depression as measured by the DSM-III and the Beck Depression Inventory in an unselected adult population. J Consult Clin Psychol 1984; 52: 892–8
- Hautzinger M, Bailer M, Worall H, Keller F. Beck-Depressions-Inventar (BDI). Huber, Bern 1994
- Beck AT. Beck depression inventory: Manual2nd ed. Harcourt Brace. 1996
- Schwarz R, Hinz A. Reference data for the quality of life questionnaire EORTC QLQ-C30 in the general German population. Eur J Cancer 2001; 37: 1345–51
- Pinquart M, Duberstein PR. Information needs and decision-making processes in older cancer patients. Crit Rev Oncol Hematol. 2004; 51: 69–80
- Repetto L, Venturino A, Vercelli M, Gianni W, Biancardi V, Casella C, et al. Performance status and comorbidity in elderly cancer patients compared with young patients with neoplasia and elderly patients without neoplastic conditions. Cancer 1998; 82: 760–5
- King MT. The interpretation of scores from the EORTC quality of life questionnaire QLQ-C30. Qual Life Res 1996; 5: 555–67
- Zimmermann PR, Poli de Figueiredo CE, Fonseca NA. Depression, anxiety and adjustment in renal replacement therapy: A quality of life assessment. Clin Nephrol 2001; 56: 387–90
- Danieli E, Airo P, Bettoni L, Cinquini M, Antonioli CM, Cavazzana I, et al. Health-related quality of life measured by the Short Form 36 (SF-36) in systemic sclerosis: Correlations with indexes of disease activity and severity, disability, and depressive symptoms. Clin Rheumatol 2005; 24: 48–54
- Lehto US, Ojanen M, Kellokumpu-Lehtinen P. Predictors of quality of life in newly diagnosed melanoma and breast cancer patients. Ann Oncol 2005; 16: 805–16
- Carlson LE, Bultz BD, Morris DG. Individualized quality of life, standardized quality of life, and distress in patients undergoing a phase I trial of the novel therapeutic Reolysin. Health Qual Life Outcomes 2005; 3: 7
- Birkhaug EJ, Aarstad HJ, Aarstad AK, Olofsson J. Relation between mood, social support and the quality of life in patients with laryngectomies. Eur Arch Otorhinolaryngol 2002; 259: 197–204