PA I N M E D I C I N E Volume 5 • Number 4 • 2004

RESEARCH ARTICLES The Longitudinal Occurrence and Impact of Comorbid Chronic Pain and Chronic Depression over Two Years in Continuing Care Retirement Community Residents Jana M. Mossey, PhD, MSN, MPH,*† and Rollin M. Gallagher, MD, MPH*‡ *Department of Epidemiology and Biostatistics, School of Public Health; †Department of Medicine, School of Medicine, Drexel University, Philadelphia, Pennsylvania; ‡Departments of Psychiatry and Anesthesiology, University of Pennsylvania School of Medicine and Pain Management Service, Philadelphia Veterans Affairs Medical Center, Philadelphia, Pennsylvania

ABSTRACT

Objective. To describe the longitudinal course of depressive symptoms and pain experienced by continuing care retirement community (CCRC) residents and to investigate the impact of comorbid chronic activity-limiting pain and chronic high depressive symptoms on physical functioning and health service use. Methods. This longitudinal study of 169 CCRC residents involved five assessments (baseline and four in-person interviews at 6-month intervals). The Geriatric Depression Scale (GDS), questions drawn from the McGill Pain Questionnaire, and self-report data on physical functioning and health care use were assessed. Individuals reporting activity-limiting pain and those with GDS scores ≥11 at three or more assessments were considered to have chronic pain or chronic depression, respectively. The analysis sample included 169 CCRC residents. Multivariate logistic regression was used to test hypotheses. Results. Pain and depressive symptoms were characterized by longitudinal stability. Of the sample, 37% met the criteria for chronic activity-limiting pain, 21% met the criteria for chronic high depressive symptoms, and 13% were comorbid. Adjusting for age and health conditions, those with chronic activity-limiting pain were five times more likely than those in the lowest pain group to persistently be in the worst two quartiles of physical functioning, as were those with even one GDS score >5. The odds of poor physical functioning were 11.2 times greater in those with comorbid chronic pain and depression. Comparable greater odds were seen in this sample for frequency of medical care visits (adjusted odds ratio [AOR] = 12.4) and consistently high use of all medical services (AOR = 3.5). Conclusions. Pain and depressive symptoms were both common and appeared remarkably stable over time. Depressive symptoms contributed significantly to the prediction of impairment associated with pain, and identification and treatment of such symptoms, even minor symptoms, could reduce pain-related impairment and health care costs in the elderly. Key Words. Chronic Depression

Pain;

Chronic

Depression;

Elderly;

Longitudinal;

Subthreshold

Reprint requests to: Jana M. Mossey, Ph.D., MSN, MPH, Drexel University School of Public Health, Mail Stop 660, 1505 Race Street, 11th Floor, Philadelphia, PA 19102-1192. Tel: (215) 762-8245; Fax: (215) 762-4088; E-mail: [email protected]. © American Academy of Pain Medicine 1526-2375/04/$15.00/335 335–348

336 Introduction

C

ross-sectional studies have demonstrated the debilitating effects of comorbid pain and depression on the functioning and well-being of older individuals [1–3]. The high prevalence of both painful conditions and depression in this population [4,5], the projected increase in the number of persons over the age of 65 [6], and the potential costs to the health care system lead to a need for clarification of the cross-sectional and longitudinal nature of this comorbid relationship and of its impact on elders. In a previous paper [1], we explored the cross-sectional relationship between pain, depression, and functional performance in independently living residents of a continuing care retirement community (CCRC). We observed comorbid subthreshold depression or DSM-III major depression and activity-limiting pain in 22.8% of the sample. For an additional 30.3%, the comorbid elevated levels of pain or depressive symptoms were less severe. While both pain and depressive symptoms were related to physical performance, the data showed that depressive symptoms, rather than pain, appeared the more influential factor. Specifically, we observed that pain, even when reported to limit activities, was related to poor physical functioning only among those individuals whose depressive symptom levels were elevated [1]. This cross-sectional analysis did not identify the directionality or causal nature of these effects— whether painful conditions that impair physical functioning cause depression or whether, in persons with pain conditions, the onset of depression causes or increases physical impairments. Both these hypotheses are supported in the literature [7,8]. However, these cross-sectional findings did highlight the importance of aggressive identification and treatment of both pain and depression in older individuals. While provocative, the cross-sectional analyses failed to provide important insights regarding the longitudinal course of pain and depression in older individuals. A review of the literature revealed a paucity of information regarding the over-time experience of elders, particularly with respect to pain levels. The several longitudinal studies on the occurrence of depressive symptoms in those over 60 years of age show subthreshold depression, which includes the minor depression category suggested for consideration in DSM-IV [9], to be substantially more prevalent than major depression [10–12]. While associated with less impairment

Mossey and Gallagher than major depression, this milder depressive symptom level appears to be strongly related to poor physical functioning and diminished life quality in those affected. Moreover, because subthreshold depression is poorly identified and treated, it appears to persist over time. Studies have shown that individuals with chronically elevated depressive symptoms experience poorer recoveries from diverse conditions such as hip fracture [13,14]. The absence of studies that include data on both pain and depression have precluded the opportunity to understand whether comorbid pain and depression persist over time and, if so, the nature of their relationship to the functioning of affected individuals. Our prior study of independently living residents in a CCRC, mentioned above [1], continued for 24 months beyond the baseline period from which the cross-sectional findings were derived. Individuals were interviewed at 6-month intervals, at which time assessments of both pain and depressive symptom levels were made. The availability of these data on a well-characterized cohort provided a unique opportunity to undertake the necessary extension of the cross-sectional studies. In this paper, we first describe the course of depressive symptoms and pain as experienced in the elderly cohort. We then investigate the occurrence of comorbid pain and depression over the two study years. Last, we examine the relationships between comorbid chronic pain and chronic depression and the cohort’s physical functioning and use of health services. We hypothesized that: longitudinal pain levels and longitudinal depressive symptom levels, alone or as comorbid conditions, have strong statistically significant associations with longitudinal physical functioning level and health service use, such that chronic activity-limiting pain and/or chronic high depressive symptoms are associated with chronic poor physical functioning, a high frequency of health care practitioner visits, and a consistently high use of different types of health care services. The study sample of individuals residing in CCRCs in the greater Philadelphia area is particularly well suited to our purposes. Residents form a homogenous group with an advantaged socioeconomic status and relatively good health status. These characteristics, along with the unrestricted access to health care afforded residents of CCRCs, substantially minimize the confounding effects of socioeconomic status and ability to pay for health services.

Comorbid Chronic Pain and Depression Among Elderly Methods

Overview of Design and Study Recruitment The data for these analyses were derived from a 24-month longitudinal study to investigate the personal and economic costs of depression in independently living elders. Details of the study methods have been reported elsewhere [1]. In brief, the study sample was drawn from the residents in the independent living sections of 25 CCRCs located in the greater Philadelphia area. Recruitment letters were sent to approximately 5,000 individuals over 60 years of age. One thousand three hundred forty-seven (1,347) residents, who disproportionately represented younger, female residents, agreed to participate in an inperson screening interview. Eligible respondents included those fluent in English, whose Blessed Dementia Scale [15] score was £6 and who appeared to have the cognitive ability to complete the study. Individuals who reported nine or more depressive symptoms on the Geriatric Depression Scale (GDS) [16] were overenrolled—one hundred seventeen such individuals agreed to participate (53% of all eligible individuals). For each of these, at least one person of the same gender and age stratum (60–74, 75–84, >84) whose GDS score was