Fax: (310) 206-5343
Partnership status influences quality of life in low-income, uninsured men with prostate cancer
Article first published online: 23 MAY 2005
Copyright © 2005 American Cancer Society
Volume 104, Issue 1, pages 191–198, 1 July 2005
How to Cite
Gore, J. L., Krupski, T., Kwan, L., Maliski, S. and Litwin, M. S. (2005), Partnership status influences quality of life in low-income, uninsured men with prostate cancer. Cancer, 104: 191–198. doi: 10.1002/cncr.21139
- Issue published online: 17 JUN 2005
- Article first published online: 23 MAY 2005
- Manuscript Accepted: 9 MAR 2005
- Manuscript Revised: 24 FEB 2005
- Manuscript Received: 4 OCT 2004
- California Department of Health Services Cancer Detection Section
- marital relationship;
- quality of life;
- medically uninsured;
- prostate cancer
Being partnered confers significant benefits in survival for patients with prostate cancer, yet little is known of the impact of relationship status on health-related quality of life (HRQOL). The authors evaluated the influence of partnership on measures of HRQOL.
The authors studied 291 patients who were enrolled in a program that provided free treatment to impoverished, uninsured men with prostate cancer. The associations between relationship status and measures of general and disease-specific HRQOL were evaluated. Results from multivariate models determined the independent effect of partnership on HRQOL.
Partnered patients were more likely than unpartnered patients to be Hispanic (58% vs. 34%) and were more likely to have elected surgical therapy (49% vs. 34%). Multivariate analyses, controlling for age, ethnicity, disease stage, and treatment type, revealed that partnered patients had better mental health (P = 0.009), less urinary bother (P = 0.011), higher spirituality (P = 0.037), and lower symptom distress (P = 0.005) than unpartnered participants.
Relationship status had a positive effect on the quality of life of low-income, uninsured men with prostate cancer. Cancer 2005. © 2005 American Cancer Society.
Gray et al. labeled prostate cancer a “relationship diagnosis.”1 Given its natural history, the myriad treatment options available, and the adverse effects of treatment on health-related quality of life (HRQOL), support systems are crucial to the care of men with prostate cancer. In fact, married men with prostate cancer may have improved survival compared with single, divorced, and widowed men.2 This improved survival may be related to compliance with screening practices, because relationship status also affects disease stage at diagnosis.3
Few investigators have evaluated the impact of relationship status on HRQOL. Most have examined differential interpretations of psychosocial distress between patient and spouse and made no comparisons between partnered and single patients.4–6 However, intuitively, it seems logical that partnered patients would have built-in support systems for coping with the psychosocial and physical obstacles that confront the patient with prostate cancer. Rose demonstrated that patients with newly diagnosed prostate cancer seek out aid from family members, modeling from friends and relatives dealing with prostate cancer, and open communication with their physicians regarding treatment and prognosis.7 Whether a patient is able to utilize those support systems likely affects HRQOL.
The objective of this study was to evaluate the influence of relationship status on HRQOL in a cohort of disadvantaged men with prostate cancer. We adjusted for clinical variables known to affect HRQOL, including age, ethnicity, disease stage, and treatment type, to ascertain the independent effect of partnership on HRQOL.
MATERIALS AND METHODS
Our study sample included all 291 low-income, uninsured men who participated in a state-funded program that provides free prostate cancer treatment. Patients with biopsy-proven prostate adenocarcinoma and a maximum annual income of 200% of the Federal Poverty Level are eligible to enroll in the program, which is called Improving Access, Counseling, and Treatment for Californians with Prostate Cancer (IMPACT). Eligibility is not restricted on the basis of disease stage, tumor grade, or a prior history of localized or systemic treatment for prostate cancer. Patients who speak a primary language other than English or Spanish are excluded from the research component of the program. Program participants receive treatment, counseling, and follow-up care for 12–18 months, including any necessary adjuvant or secondary therapies, all at no charge. Details of the IMPACT program have been published previously.8
Clinical information obtained at enrollment included demographic data as well as an assessment of comorbid conditions; tumor characteristics, including stage and grade; and the type of prior treatments received, if any. Participants defined their relationship status as either “living with spouse or partner,” “in a significant relationship, but not living together,” or “not in a significant relationship.” We classified the first two responses as partnered and the final response as single or unpartnered. We used the Charlson index to convert comorbidity into a numeric value, with a higher number indicating a greater burden of chronic comorbid diseases.9 We categorized disease stage according to the American Joint Committee on Cancer/International Union Against Cancer Tumor-Lymph Node-Metastases (TNM) classification and stage groupings,10, 11 as follows: localized (prostate adenocarcinoma Stage I–II [T ≤ 2b,N0,M0]: tumor localized to the prostate gland [T ≤ 2b], no regional lymph node metastasis [N0], no distant metastasis [M0]), regional (prostate adenocarcinoma Stage III [T ≥ 3,N0,M0]: tumor extending beyond the prostatic capsule [T ≥ 3], no regional lymph node metastasis [N0], no distant metastasis [M0]), and metastatic (prostate adenocarcinoma Stage IV [T1–T4,N1,M0–1; T1–T4,N0–N1,M1]: any tumor with regional lymph node involvement [N1] or distant metastasis [M1]). Treatment type reflected the primary treatment received, because IMPACT enrollees were not reclassified on the basis of adjuvant therapies unless they were managed initially with watchful waiting.
Participants in the evaluation component of IMPACT completed the Men's Health Research Survey, a battery of validated measures of HRQOL that also assesses socioeconomic and demographic variables. We obtained informed consent from each participant, and all recruitment and research protocols were approved by the University of California–Los Angeles (UCLA) Institutional Review Board and were compliant with the Health Insurance Portability and Accountability Act of 1996. We evaluated general HRQOL with the RAND Medical Outcomes Study Short Form 12 version 2 (SF-12)12, 13 and the complete Mental Health Inventory (MHI-5) from the RAND 36-Item Health Survey (SF-36).14 The SF-12 abbreviates the SF-36 and measures both physical and mental health, with composite scores (PCS and MCS, respectively) normalized to the general population. A score of 50 reflects the normative mean with a standard deviation of 10. The PCS and MCS correlate well with the SF-36 summary scales (R2 = 0.938 and 0.905, respectively) and have high test-retest reliability (r = 0.89 and 0.76, respectively).12 The MHI-5 comprises all 5 items from the SF-36 that measure emotional well being. Responses are scored from 0 to 100 and then averaged, with higher scores indicating better HRQOL. We assessed disease-specific quality of life with the UCLA Prostate Cancer Index Short Form (PCI-SF),15 which is a validated short form of the PCI and measures urinary, sexual, and bowel habits in two domains: function and bother. Similarly, responses are scored from 0 to 100, with higher scores indicating better outcome. The PCI-SF is both valid and reliable in men with and without prostate cancer.16
Other instruments in the Men's Health Research Survey measured symptom distress, anxiety, spirituality, and the patient's confidence in their patient-physician interactions. We evaluated the burden of cancer-related symptoms with McCorkle and Young's Symptom Distress Scale (SDS).17 Questions assess the amount of symptom distress patients experience related to nausea, mood, appetite, insomnia, pain, mobility, fatigue, bowel pattern, concentration, and appearance. Responses are scored from 1 to 5, with higher scores indicating greater distress. The SDS demonstrates high reliability (reliability coefficient α = 0.82).17 We measured the level of anxiety related to a fear of cancer recurrence with a five-item subscale of the Memorial Anxiety Scale for Prostate Cancer,18 which scores responses from 1 to 5, with a higher score indicating a lower level of anxiety related to fear of recurrence. The Fear of Recurrence subscale has performed well with high reliability (reliability coefficient α = 0.76).18 We assessed spirituality with the Functional Assessment of Chronic Illness Therapy Scales Spirituality subscale (FACIT-Sp).19 The FACIT-Sp is a 12-item survey with a Meaning/Peace component, which assesses participants' sense of purpose in life, and a Faith component, which quantifies the level of comfort and strength patients derive from their spiritual beliefs. Responses are scored from 0 to 4 and then summed, with higher scores indicating a higher level of spirituality. The FACIT-Sp has been validated extensively and demonstrates high reliability (reliability coefficient α = 0.87).19 The Perceived Efficacy in Patient-Physician Interactions (PEPPI) scale is used to assess participants' confidence level in their interactions with their physician.20 The PEPPI is a 5-item scale that evaluates patients' confidence in their ability to have informative and beneficial discussions with their physicians. Responses are scored from 1 to 5 and then summed, with a higher score reflecting a lower level of confidence in their patient-physician interactions. The PEPPI demonstrated high reliability (reliability coefficient α = 0.93) and has proved valid.20 Spanish-language versions of the instruments included in the Men's Health Research Survey have been validated in Hispanic populations.13, 21, 22
Descriptive statistics are presented for demographic and clinical data, comparing partnered and unpartnered patients using chi-square analysis for categorical variables and the two-sample t test for continuous variables. Univariate analyses were performed to evaluate the relation between partnership status and measures of general and disease-specific HRQOL. Multivariate models were created based on variables with P values ≤ 0.2 in univariate analysis to determine the independent effect of partnership on HRQOL. We also included age at enrollment, ethnicity, disease stage, and primary treatment type as covariates in our multivariate models. For HRQOL measures that were associated significantly with partnership status on multivariate modeling, we performed hierarchical modeling to evaluate the contribution of partnership status to the model. The first block included age, race, tumor stage, and treatment; and the second block included partnership status. To evaluate whether the effect of relationship status on HRQOL was modified by other patient characteristics, we performed a separate analysis in which we incorporated interaction terms into the final multivariate models. We examined the interaction between relationship status and both age and ethnicity for all HRQOL outcomes and between relationship status and both urinary and sexual function for urinary and sexual bother outcomes, respectively. All statistical analyses were performed with SAS software (version 8.02; SAS Institute, Cary, NC).
Complete demographic and clinical data were available for 291 participants at enrollment. Table 1 compares these data between partnered and unpartnered participants. Partnered patients were more likely to be both Hispanic and less educated. Moreover, partnered patients were more likely to have undergone surgery as primary treatment and were less likely to opt for watchful waiting.
|Characteristic||No. of patients (%)||P value|
|Partnered patients (n = 211)||Unpartnered patients (n = 80)|
|Mean age (yrs) ± SD||61.9 ± 6.7||61.4 ± 7.9||0.548|
|White||40 (19.0)||27 (33.7)||0.0007|
|African American||31 (14.7)||21 (26.3)|
|Hispanic||123 (58.3)||27 (33.7)|
|Other||17 (8.0)||5 (6.3)|
|< High school||96 (45.5)||23 (28.8)||0.014|
|High school||91 (43.1)||40 (50.0)|
|≥ College||24 (11.4)||17 (21.2)|
|Comorbidity index (n = 270 men)|
|0||102 (53.1)||45 (57.7)||0.495|
|≥ 1||90 (46.9)||33 (42.3)|
|Primary treatment (n = 285 men)|
|Radical prostatectomy||101 (49.3)||27 (33.8)||0.020|
|Radiation||57 (27.8)||21 (26.3)|
|Watchful waiting||6 (2.9)||6 (7.5)|
|Hormone therapy||41 (20.0)||26 (32.5)|
|Clinical stage (n = 286 men)|
|Localized||127 (61.6)||42 (52.5)||0.283|
|Localized regional||28 (13.6)||11 (13.7)|
|Metastatic||51 (24.8)||27 (33.8)|
Table 2 displays univariate analyses of the association between relationship status and measures of HRQOL. Partnered patients had better mental health, less urinary bother, lower symptom distress, higher spirituality, and less confidence in patient-physician efficacy.
|Measure||Mean ± SD score||Effect size||P value|
|Partnered patients||Unpartnered Patients|
|Mental Composite score||42.3 ± 7.6||41.8 ± 8.3||−0.067||0.630|
|Physical Composite score||45.7 ± 11.9||43.1 ± 12.9||−0.219||0.115|
|Mental Health Inventory||70 ± 22||62 ± 25||−0.339||0.014|
|Prostate Cancer Index Short Form|
|Sexual function||31 ± 31||34 ± 32||0.098||0.473|
|Sexual bother||31 ± 37||33 ± 38||0.035||0.798|
|Urinary function||72 ± 30||67 ± 30||−0.149||0.259|
|Urinary bother||70 ± 34||60 ± 38||−0.284||0.038|
|Bowel function||75 ± 26||71 ± 27||−0.145||0.277|
|Bowel bother||72 ± 31||68 ± 36||−0.125||0.366|
|McCorkle Symptom Distress Scale||20 ± 7||23 ± 8||0.395||0.0038|
|Anxietya||16 ± 5||15 ± 5||−0.244||0.071|
|Spirituality (per point on the FACIT-Sp)||39 ± 8||36 ± 10||−0.517||0.0032|
|PEPPI||21 ± 4||20 ± 5||−0.279||0.039|
Table 3 presents the results of multivariate regression analyses, which controlled simultaneously for age, ethnicity, disease stage, and primary treatment type. Relationship status had significant, positive, independent effects on mental health, urinary bother, symptom distress, and spirituality. Among the variables that were significant in univariate analyses, all except patient-physician self-efficacy were significant in the multivariate models. Results of the hierarchical modeling demonstrated the additional variance accounted for by partnership status for mental health (R2 = 0.075 vs. 0.054), urinary bother (R2 = 0.038 vs. 0.019), symptom distress (R2 = 0.092 vs. 0.070), and spirituality (R2 = 0.162 versus 0.082).
|Predictor||SF-12 PCS||MHI-5||PCI-SF: Urinary bother||FACIT-Sp||SDS||Anxietya||PEPPI|
|Effect||P value||Effect||P value||Effect||P value||Effect||P value||Effect||P value||Effect||P value||Effect||P value|
|Relationship status vs. single)||2.47||0.136||8.17||0.009||12.0||0.011||2.27||0.037||−2.82||0.005||0.82||0.160||1.02||0.101|
|Ethnicity (vs. white)|
|Stage (vs. localized)|
|Primary treatment (vs. surgery)|
Our multivariate model also identified significant associations between covariates and HRQOL outcome measures. Age was associated independently with improved mental health. Compared with white men, African-American men had higher spirituality and lower anxiety related to fear of recurrence. Hispanic men had worse physical functioning, worse mental health, and higher spirituality than white men. Patients who had metastatic disease had worse physical functioning, more symptom distress, higher anxiety related to disease progression, and lower spirituality compared with patients who had localized prostate cancer. No interaction terms were significant.
The current study has produced several important findings. Regardless of age, ethnicity, disease stage, or treatment type, being in a marital-type relationship was associated with significantly better quality of life. Mental health scores in partnered patients were fully eight points higher than the scores of single men, with unpartnered patients registering a mean MHI-5 score of 62. By comparison, mean MHI-5 scores in patients with the chronic debilitating conditions of diabetes mellitus, congestive heart failure, and chronic pulmonary disease range from 68 to 77.14 The salutary effect of relationship status on mental health was as sizeable as the negative effect of advanced tumor stage on mental health. In patients with other cancers, social support systems have reduced emotional distress and promoted adaptation to the cancer diagnosis.23 Men with cancer seem to benefit from spousal support more than women with cancer. Among patients with breast and prostate cancers, female partners have a better understanding of their husbands' experience with prostate cancer than male partners of women with breast cancer.6 Older patients and those in longer relationships seem to benefit more.6 Older patients may have reduced anxiety related to their cancer diagnosis, an effect that may be mediated by a reduced expectation for prolonged survival and a greater experience with chronic diseases. Prior investigations corroborating the association between older age and better mental health did not extend theories to explain this emotional benefit.24, 25
Partnered patients in our study also had significantly less urinary bother than unpartnered participants. Independent of treatment type, partnered patients reported urinary bother scores that were 12 points better than those reported by unpartnered patients. Partnered patients may benefit from their spouses' experience with incontinence, which is an extremely prevalent condition among women.26 Phillips et al. showed that, even in men with severe postprostatectomy incontinence, spousal support allowed patients to cope with frustration secondary to urinary incontinence.27 Partnered patients gained perspective on their urinary problems, because most minimized the importance of their poor urinary function relative to the benefit of the prolongation of life gained through surgery.27, 28 Although emotional support facilitates improved perceptions of distress from urinary incontinence, the influence of a partner on urinary bother also can extend to the physical and financial realms. Support with the purchase of pads and other incontinence aids may factor into differences in experienced bother, although this may not have been as relevant in our study sample, which received free pads and incontinence aids. Previous studies examining the impact of sociodemographic variables on HRQOL showed that married men had better baseline and posttreatment urinary function than unmarried men.29, 30 In our cohort, urinary function scores were comparable between partnered and unpartnered men. Intuitively, it seems logical that having a partner more likely would benefit bother, which is a subjective measure of distress, rather than function, which is an objective measure of the frequency and severity of incontinent episodes.
Although urinary bother differed significantly between partnered and single participants, no difference was noted in either sexual function or sexual bother. Although spousal support ameliorates distress from urinary difficulty, partnered men are bothered just as much as single men by sexual dysfunction. This may be explained by our patients' unwillingness to rely on support systems as a coping mechanism for managing sexual dysfunction. Previous studies examining interactions between patients with prostate cancer and their spouses found that, although men were open about problems like physical disability and fatigue, they often did not share their prostate-related health problems, especially those related to sexual function.31, 32 Even when open marital discussion occurs, spousal reassurance, although valuable, does little to assuage patient concerns about sexual function.27
Whereas disease-specific symptoms varied little between partnered and single patients, with the exception of urinary bother, distress from general cancer-related symptoms was significantly lower in partnered men, although both groups had relatively low levels of symptom distress. By comparison, mean SDS scores for patients with advanced lung cancer, breast cancer, and coronary artery disease ranged from 17.8 to 26.7.33, 34 Few investigators have examined the impact of partnership on general symptom distress, although one study of patients with breast cancer failed to show a correlation.4 The significant benefit seen in our partnered patients may reflect the tendency for men with prostate cancer to be forthcoming with physical symptomatology, especially fatigue. The impact of reduced symptom distress is substantial: It has been correlated highly with both HRQOL and survival.35–38 The structure of the McCorkle SDS may have confounded interpretation of the results, especially in our study sample. The instrument measures the severity of the symptoms but does not differentiate whether these effects are treatment-related or disease-related. Although our partnered patients did have lower symptom distress, partnered and unpartnered study patients differed significantly in primary treatment type and disease stage at baseline.
Finally, the level of spirituality was higher in partnered patients than in single patients. Relationship status likely influences spirituality, which, in turn, may be associated with HRQOL. A pilot study of men with prostate cancer focused on the use of faith as a coping mechanism, revealing that a significant proportion of patients with prostate cancer may want faith to factor into their management.39 Brady et al. demonstrated that spirituality positively affected physical and emotional well being in a cohort of patients with acquired immunodeficiency syndrome and a variety of cancer diagnoses.40 The same likely is true for patients with prostate cancer.
Understanding the relation of partnership status to quality of life in patients with prostate cancer can help guide strategies that may benefit both partnered and unpartnered men. Prostate cancer support groups almost universally are available and provide patients with an opportunity to interact with others who have had similar experiences, both physically and emotionally. Previous studies have shown that patients with prostate cancer in support groups have better disease-specific quality of life, independent of their marital status.41 Despite the proven benefit and availability of cancer support groups, male participation is poor. In a surveyed sample of patients with prostate cancer, only 13% participated in support group meetings, compared with 33% of women with breast cancer.42 In addition, patients who actually attended support groups tended to be better educated and more affluent than the general population of men with prostate cancer. Whether time constraints or a lack of counseling limits access to support groups among low-income patients, the general population of men with prostate cancer needs more information regarding the benefit and availability of prostate cancer support groups.
This study had several limitations. First, although we had information related to the partnership status of our patients, we had no information regarding the quality of the relationship. Studies have shown that emotionally healthy couples have decreased psychological distress, a phenomenon termed dyadic strength.5 Moreover, in individuals with poor marital adjustment, the stress of a cancer diagnosis in partnered patients actually can exacerbate depressive and anxiety symptoms.43 That is, the partnership actually may negate the additional support of being in a relationship. In not controlling for dyadic adjustment, our results may have underestimated the effect of a relationship on HRQOL. Similarly, we had little information on the social matrix of our patients. Being partnered likely is associated with unmeasured variables that influenced our outcomes of interest more directly, such as emotional support and family structure.
Second, the current study was limited by the small size of the study sample, which comprised a heterogeneous combination of men with different disease stages and primary treatment types. We controlled for these variables in our models; however, effects may persist. We sought to investigate the association between relationship status and HRQOL in all men with prostate cancer. Therefore, we included patients in all strata of prostate cancer care. The incorporation of different disease stages and treatment types may make some of our findings more meaningful. Partnered men tended to have lower disease stage and more commonly underwent radical prostatectomy as primary therapy. Despite these factors predisposing to higher urinary bother, our partnered patients had urinary bother scores that were 10 points better than the scores of single participants. Controlling for disease stage and treatment in our multivariate model only exacerbated that difference. Furthermore, because men are enrolled regardless of disease stage or a history of prior treatments for prostate cancer, information on time since treatment largely is unavailable for our cohort. This may have had an impact on our results, because HRQOL scores can vary along the recovery trajectory.
Third, the men in our sample were heterogeneous with respect to ethnicity, and this cultural diversity may have introduced bias in the reporting of the HRQOL outcomes studied. Although the instruments that comprise the Men's Health Research Survey have been validated extensively in various language and ethnic groups, comprehension and acculturation may have confounded the interpretation of our results. A significant proportion of our patients had less than a high school level education. We attempted to overcome this limitation through telephone administration of surveys by fluent interpreters; however, literacy and comprehension still may have influenced our results. Furthermore, investigators have shown that ethnic minorities may report comparable health states differentially on HRQOL instruments.44 Differential reporting of the severity of various health states may have produced ethnic differences in HRQOL domains in which the true differences were minimal. Further work will be needed to address issues of acculturation in the measurement of HRQOL; because, currently, culture-specific instruments for general and disease-specific HRQOL in men with prostate cancer are not available.
Finally, our sample was a cohort of severely impoverished men. Coping strategies that benefit health, including optimism, a sense of personal control, and social support, are less available to patients of low socioeconomic status.45 Therefore, single participants in our study may represent a particularly isolated cohort of men with prostate cancer, and our results may not be generalizable to others.
Despite these limitations, we found an association between relationship status and quality of life in men with prostate cancer. Both general and disease-specific facets of HRQOL were affected positively in partnered patients. Clinicians who care for patients with prostate cancer need to address coping and social support mechanisms to encourage the beneficial aspects of partnership and to overcome the detrimental effects of being single.
- 10SobinLH, WittekindC, editors. TNM classification of malignant tumors, 5th ed. New York: John Wiley & Sons, 1997.
- 11FlemingID, CooperJS, HensenDE, et al., editors. AJCC cancer staging manual, 5th ed. Philadelphia: Lippincott-Raven, 1998.
- 13How to score version 2 of the SF-12 Health Survey. Boston: Quality Metric Incorporated and Health Assessment Laboratory, 2002.
- 14SF-36 Health Survey: manual and interpretation guide. Boston: The Health Institute, New England Medical Center, 1993.
- 24Mental health in men treated for early stage prostate carcinoma: a posttreatment, longitudinal quality of life analysis from the Cancer of the Prostate Strategic Urologic Research Endeavor. Cancer. 2002; 95: 54–60., , , , .