Pain Severity in Diabetic Peripheral Neuropathy is Associated with Patient Functioning, Symptom Levels of Anxiety and Depression, and Sleep

doi:10.1016/j.jpainsymman.2005.04.009374 Journal of Pain and Symptom Management Vol. 30 No. 4 October 2005
Original Article
Pain Severity in Diabetic Peripheral Neuropathy is Associated with Patient Functioning, Symptom Levels of Anxiety and Depression, and Sleep Mugdha Gore, BPharm, PhD, Nancy A. Brandenburg, PhD, Ellen Dukes, PhD, Deborah L. Hoffman, PhD, Kei-Sing Tai, MS, and Brett Stacey, MD Avalon Health Solutions, Inc. (M.G., D.L.H., K.-S.T.), Philadelphia, Pennsylvania; Pfizer, Inc.
(N.A.B., E.D.), New York, New York; and Pain Management Center (B.S.), Oregon Health
and Science University, Portland, Oregon, USA
Abstract Our goal was to evaluate pain severity, pain-related interference with function, sleep impairment, symptom levels of anxiety and depression, and quality of life among patients with painful diabetic peripheral neuropathy (DPN). Participants in a burden of illness survey (n5
255) completed the modified Brief Pain Inventory-DPN (BPI-DPN), MOS Sleep Scale, Hospital Anxiety and Depression Scale (HADS), Short Form Health Survey-12v2 (SF-12v2), and the EuroQoL (EQ-5D). Patients were 61 6 12.8 years old (51.4% female), had diabetes for 12 6 10.3 years and painful DPN for 6.4 6 6.4 years. Average and Worst Pain scores (BPI-DPN, 0--10 scales) were 5.06 2.5 and 5.66 2.8. Pain substantially interfered ($4 on 0--10 scales) with walking ability, normal work, sleep, enjoyment of life, mood, and general activity. Moderate to severe symptom levels of anxiety and depression (HADS-A and HADS-D scores $11 on 0--21 scales) occurred in 35% and 28% of patients, respectively. Patients reported greater sleep problems compared with the general U.S. population and significant impairment in both physical and mental functioning (SF-12v2) compared with subjects with diabetes. The mean EQ-5D utility score was 0.5 6 0.3. Greater pain levels in DPN (mild to moderate to severe) corresponded with higher symptom levels of anxiety and depression, more sleep problems, and lower utility ratings and physical and mental functioning, (all Ps ! 0.01). Painful DPN is associated with decrements in many aspects of patients’ lives: physical and emotional functioning, affective symptoms, and sleep problems. The negative impact is higher in patients with greater pain severity. J Pain SymptomManage 2005;30:374--385. 2005 U.S. Cancer Pain Relief Committee. Published by Elsevier Inc. All rights reserved.
Key Words Painful diabetic peripheral neuropathy, pain severity, sleep impairment, symptoms levels, anxiety, depression
Address reprint requests to: Mugdha Gore, BPharm, PhD, Avalon Health Solutions, Inc., 1528 Walnut Street, Suite 1507, Philadelphia, PA 19102, USA.
Accepted for publication: April 13, 2005.
2005 U.S. Cancer Pain Relief Committee Published by Elsevier Inc. All rights reserved.
Introduction Diabetic neuropathies are a family of nerve
disorders caused by diabetes mellitus. Nearly
0885-3924/05/$--see front matter doi:10.1016/j.jpainsymman.2005.04.009
Vol. 30 No. 4 October 2005 375Triad of Pain, Sleep, Anxiety/Depression in DPN
45% of diabetes patients develop neuropathy during the course of their disease.1 Distal symmetric sensorimotor polyneuropathy or diabetic peripheral neuropathy (DPN) is the most common of all diabetic neuropathies.2--4
In a cross-sectional survey at the Mayo clinic, 47.3% of all diabetes patients or 78% of those with any type of diabetic neuropathy had DPN.2 The prevalence of clinician-diagnosed symptomatic or painful DPN was 13--15%.2
Many patients with DPN experience neuro- pathic pain, typically characterized as burning, tingling, electric, sharp, shooting, and lanci- nating, which initially starts in both feet and may progress to involve calves, fingers, and hands (stocking and glove pattern).5 Risk factors associated with DPN include worsening glucose tolerance, older age, longer diabetes duration, drinking alcohol, and cigarette smoking.4,6--9
There is no cure for diabetic neuropathy. Treatment approaches include decelerating the progressive loss of nerve function through maintenance of glycemic control and pain management. Select antidepressants, anti-epi- leptics, and opioids were recently recommen- ded as first-line treatments for neuropathic pain.10 However, dose-limiting side effects often reduce the effectiveness of these medi- cations and leave DPN patients in continuing pain.
Pain associated with diabetic neuropathy substantially affects the quality of life among patients with this condition.5,11,12 Patients with painful DPN report substantial interference (scores $4 on 0--10 scales, where 0 5 does not interfere and 10 5 completely interferes) in sleep, enjoyment of life, recreational activities, normal work, mobility, general activity, social activities, and mood;5 and greater impairment in energy, sleep, physical mobility, and emo- tional reactions when compared to diabetic and non-diabetic controls.11
Chronic pain is a subjective experience that affects cognitive and emotional dimensions, impairs mood and thinking,13,14 causes func- tional restrictions and often inhibits daily activities and work.15--18 Neuropathic pain is particularly distressing, as it is associated with a high degree of suffering and does not generally decline over time.19 Chronic pain, sleep disturbance, and affective disorders often occur simultaneously20--23 and are purported to
have multi-directional relations with each other. Approximately 50--70% of individuals attending chronic pain clinics report sleep impairment,24--26 over 20% have major depres- sion,27--30 and epidemiological studies suggest that these patients have higher levels of anxiety than populations without pain.31--33 Further, studies in specific pain conditions report a positive association between higher pain intensity and lower levels of sleep quality (insomnia, middle of night awakening, and restless sleep34,35). Patients with comorbid de- pression and chronic pain also tend to report higher levels of pain, greater interference due to pain, tend to be less active, and report greater disability.36--38 In addition, pain intensity and depressive symptoms have been found to have a significant association in a majority of studies evaluating these constructs.29
Despite the prevalence of DPN, and the acknowledgment that pain, a common prob- lem in DPN patients, can reduce quality of life, a comprehensive assessment of the psychoso- cial burden among patients with painful DPN is lacking. Several previous studies have re- ported on the influence of diabetic peripheral nerve involvement39,40 or the effect of in- creasing neuropathy severity on quality of life;41 however, no information on the pres- ence or absence of pain among these patients is provided.
The two previous studies5,11 that reported on the patient level impact of painful DPN limited their assessments to description of pain and quality-of-life evaluations. Although these stud- ies provided useful information on a condition where literature on patient outcomes was other- wise sparse, several author-acknowledged limi- tations reduce the generalizability of their results. Sample sizes for both studies were relatively small (n 5 41;11 and n 5 1055) and the study samples were not geographically5,11 or ethnically diverse (91% White5). The Benbow and colleagues study11 recruited patients from only one diabetes clinic. Further, the Galer et al. study5 recruited subjects fromamong thosewho had inquired about participating in clinical trials for painful diabetic neuropathy, and were thus self-selected for the symptom of pain and potentially those with more severe pain. Al- though both studies acknowledged sleep and affective distress to be important, no formal assessments of either were performed, limited
376 Vol. 30 No. 4 October 2005Gore et al.
information on medical comorbidities was obtained, and all data (including medical history) were patient-reported. Lastly, because the psychometric properties of a modified version of the Brief Pain Inventory (BPI),42
the primary pain measure in one study,5 were not reported, its validity for the characterization of DPN pain could not be assumed.
Without a complete understanding of the multidimensional human burden among DPN patients, treatment of this painful condition is likely to remain unsatisfactory. Thus, our goal in the present study was to address some of the limitations of previous studies by investigating all relevant aspects of the humanistic burden among patients with painful DPN (pain severity, pain-related interference with func- tion, sleep impairment, symptom levels of anxiety and depression, functional status, and quality of life) in a large, geographically diverse sample of patients treated in commu- nity-based settings. Because pain, sleep impair- ment, and symptoms of anxiety and depression occur simultaneously in other chronic pain conditions, we also investigated the interrela- tionship between these variables among pa- tients with painful DPN. Lastly, we evaluated the psychometric properties of the BPI, mod- ified to characterize pain experience among patients with DPN (BPI-DPN). These results are reported in a separate paper.43
Methods Sample Selection
Primary care physicians, endocrinologists, anesthesiologists, and neurologists from set- tings across the United States (West, South- west, Midwest, Southeast, and Northeast) recruited patients into this community-based cross-sectional survey. Patients were invited to participate as they presented for a regularly scheduled visit. Study eligibility criteria were as follows: female or male patients $18 years of age; able to provide informed consent in- cluding permission to obtain information from patients’ medical records; able to read and understand English; present in the re- cruiting physician’s practice for at least three months prior to the survey; and have physi- cian-diagnosed diabetic distal symmetrical sensorimotor polyneuropathy with painful
symptoms (burning, prickling, tingling, and/ or shooting pain in toes, feet, legs, and/or hands) of at least three months’ duration. All study materials, including the informed con- sent forms, were approved by an institutional review board and were in compliance with the Health Insurance Portability and Account- ability Act. Patients were excluded from the study if
they had a serious or unstable medical or psychological condition that, in the opinion of the physician, would compromise participa- tion in the survey; had neurologic disorders unrelated to diabetic peripheral neuropathy; had any other pain condition; or had skin conditions that were likely to confound the assessment of diabetic peripheral neuropathic pain. Additionally, because our goal was to assess DPN pain in usual care settings, to ensure that pain and other study assessments (e.g., functional status and quality of life) would not be confounded by unknown efficacy and side effects of new investigational drugs, we excluded patients who had participated or were planning participation in an investiga- tional drug study within 90 days prior to the survey, or during the survey. Patient eligibility was assessed by study physicians. Study nurses/ physicians also reviewed the informed consent form and the study survey with patients and addressed any questions. A total of 265 patients met study eligibility
criteria and were recruited between April and October 2003. All of the 265 eligible patients agreed to participate in the study, were given the mail-in survey, and were offered a U.S.$ 50 incentive for completing the survey. Study nurses/physicians completed a case report form documenting medical and surgical histo- ries of patients (obtained from patient charts) who returned completed surveys.
Measures Basic demographic data and DPN history
information, such as years with diabetes, years since onset of painful DPN, and years of prescription medication use for painful DPN, were obtained from patients. Additionally, information on type of diabetes, use of insulin, and presence of select pain and other chronic comorbidities was obtained from physicians (patient chart data).
Vol. 30 No. 4 October 2005 377Triad of Pain, Sleep, Anxiety/Depression in DPN
The primary pain measure in this study was a short-form of the BPI modified to character- ize pain experience among patients with DPN (BPI-DPN). The BPI is a patient-rated in- strument that measures severity of pain (aver- age, worst, least, pain now), on 0 to 10 scales (0 5 no pain and 10 5 pain as bad as you can imagine) and its interference with seven functional areas, general activity, mood, walk- ing ability, normal work, relations with others, and enjoyment of life using 0 to 10 interfer- ence scales (0 5 does not interfere and 10 5
completely interferes). Responses to these items are used to generate a pain severity score and a pain interference score. Recall for all items is past 24 hours. The BPI was chosen for this study because it has a numeric rating scale format that is recommended for the measurement of neuropathic pain,10,44 it is widely used in other pain studies, and has proven to be reliable and valid.42 Cronbach’s alpha values for the BPI scales are high (O0.85) in most studies. Modification of the BPI included addition of the words ‘‘due to your diabetes’’ to each item (e.g., please rate your pain due to your diabetes). Patients with painful DPN tend to be older and suffer from other comorbid pain conditions. Thus, this modification was considered important to distinguish between pain due to DPN and pain due to other causes.
The impact of painful DPN on self-reported sleep quality and quantity over the preceding four weeks was assessed using the Medical Outcomes Study (MOS) Sleep Scale.45 The MOS-Sleep Scale is a 12-item measure used to assess the key constructs of sleep. Nine of the twelve items can be combined to derive a Sleep Problems Index (SLP9). Sleep attributes cap- tured by this index include difficulties with falling asleep, staying asleep, daytime sleepi- ness, and feelings of being rested. The scale has been found to be reliable and valid with good overall measurement properties. Internal consistency reliability estimates for the MOS Sleep Scale in the general U.S. population were between 0.75 and 0.86.46
Depression among chronic pain patients has been assessed using clinical diagnostic criteria and a variety of self-report measures. The caveat regarding the use of self-report mea- sures with chronic pain patients is that such measures may overestimate the prevalence of
depression. Nonetheless, self-report measures of depression have been found to be highly reliable and valid in chronic pain popula- tions.47--50 Furthermore, high scores on self- report measures may also be associated with disability.51--53 Thus, the Hospital Anxiety and Depression Scale (HADS),54 a brief, self- administered questionnaire, with demonstrat- ed reliability and validity,55--57was used to evaluate the impact of painful DPN on affective distress (symptoms of anxiety and depression).
The HADS consists of two subscales, one measuring anxiety (HADS-A) and the other measuring depression (HADS-D). Each sub- scale comprises 7 items, and subjects assess how each item applies to them on a scale ranging from no feelings of anxiety or de- pression (0) to severe feelings of anxiety or depression (3). The anxiety subscale analyzes a state of generalized anxiety (including anxious mood, restlessness, anxious thoughts, panic attacks), whereas the depression sub- scale focuses on a state of lost interest and diminished pleasure response (‘‘lowering of hedonic tone’’). Whereas cognitive aspects are emphasized, somatic symptoms of anxiety and depression are deliberately not covered so as to increase the likelihood that the HADS measures affective disorder rather than symp- toms of pain or other illness. The HADS has been found to be reliable in several studies, with Cronbach’s alpha values for the HADS-A subscale ranging from 0.68 to 0.93, and those for the HADS-D subscale ranging from 0.67 to 0.90.55
The impact of painful DPN on patient quality of life was assessed using the Short- Form 12 version 2 (SF-12v2) health survey.58
The SF-12v2 consists of 12 items selected from the SF-36,59 a widely used survey in clinical practice, research, and health policy evalua- tions. Responses to the 12 items of the SF-12v2 are used to generate a health profile compris- ing eight scales and two summary measures. The Physical Component Summary measure (PCS) incorporates the Physical Functioning, Role-Physical, Bodily Pain, and General Health scales; the Mental Component Summary mea- sure (MCS) incorporates the Vitality, Social Functioning, Role-Emotional, and Mental Health scales. The SF-12v2 has a 4-week recall period and has good reliability and validity.58
378 Vol. 30 No. 4 October 2005Gore et al.
Cronbach’s alpha values for the SF-12v2 scales in the general U.S. population were between 0.73 and 0.89.58
The EuroQoL (EQ-5D) was used to evaluate patients’ perceived health status.60 The EQ-5D is a 5-item generic health status measure that assesses the extent of impairment (no prob- lems, some problems, and extreme problems) in five aspects of patient well being: mobility, self-care, usual activities, pain discomfort, and anxiety/depression.60 Responses to individual items are used to calculate a utility score. The validity and reliability of the EQ-5D has been established in a number of disease states.
Lastly, two single-item, 0 to 100 scales were used to determine patients’ assessments of current health status (0 5 worst possible health and 100 5 perfect health) and health status given a hypothetical complete relief of DPN pain.
Information on healthcare utilization (pre- scription and over-the-counter medications, consults [visits and telephone calls] with healthcare professionals, hospitalizations, di- agnostic, and surgical procedures), medication satisfaction, and work productivity and activity limitations were obtained from patients and physicians. These findings are being reported in a separate paper.
Data Analyses Descriptive statistics were used to describe
the study sample, means and standard devia- tions for continuous variables, and numbers and percents for categorical variables. Study instruments (BPI-DPN, MOS Sleep Scale, HADS, SF-12v2, EQ-5D) were scored using standard scoring conventions. A score ranging from 0 to 100 was obtained for the 9-item MOS Sleep Problems Index (SLP9), with higher scores indicating greater sleep problems;45
a score ranging from 0 to 21 was obtained for both HADS-A and HADS-D subscales, with scale scores between 0 and 7 considered as ‘‘normal,’’ those between 8 and 10 as ‘‘mild,’’ those between 11 and 14 as ‘‘moderate,’’ and those between 15 and 21 as ‘‘severe’’ for both symptom levels of anxiety and depression;54
normalized scores (0--100 scales scores linearly transformed to have a mean of 50 and standard deviation of 10 in the 1998 general U.S. population) were obtained for each of the SF-12v2 scales, with higher scores reflecting
better quality of life.58 Quality of life of the study sample (SF-12v2) was compared to normative data from the general U.S. popula- tion and U.S. patients with diabetes.58 A utility score ranging from 20.6 to 1.0 was derived from responses to items on the EQ-5D. Higher utility scores indicate better health status.60
The internal consistency reliability of study measures was assessed using Cronbach’s alpha values. The study sample was divided into three groups based on pain severity categories (mild, moderate, severe) derived for patients with painful DPN.61 Patients who rated their pain between 0 and 3 on the BPI-DPN average pain item were assigned to the ‘‘mild’’ pain group, those with ratings between 4 and 6 were assigned to the ‘‘moderate’’ pain group, and those with ratings between 7 and 10 were assigned to the ‘‘severe’’ pain group. Analyses of variance and F-tests were used to determine if patient ratings on the MOS Sleep Scale, HADS, SF-12v2, EQ-5D, and the global health scale differed based on levels of pain severity. No corrections were made for multiplicity. All analyses were conducted using the SAS soft- ware system, PC version 8.0.
Results A total of 265 patients met study eligibility
criteria and all agreed to participate in the study. Of these, 255 patients returned com- pleted surveys resulting in a response rate of 96.2%. Demographic and clinical character- istics of the study sample are described in Table 1. On average, patients were 61.3 (SD 5
12.8) years old, 51.4% were women, and subjects had diverse ethnic backgrounds: Caucasian (53.3%), African American (18.8%), Latino(9.8%),multiracial (3.1%),Asian(2.4%) and Native American (0.4%). Less than a third were employed either part-time (!29 hours per week) or full-time ($30 hours per week) and over two-thirds (68.6%) had completed at least a high school degree. Patients had diabetes for an average of 12 years (SD 5
10.3), painful DPN an average of 6.4 years (SD 5 6.4), and had used prescription medications for painful DPN an average of 5.4 (SD 5 6.3) years. Physician-reported data indicated that a majority (86.3%) of patients had Type II diabetes and 40.9% were taking insulin; nearly
Vol. 30 No. 4 October 2005 379Triad of Pain, Sleep, Anxiety/Depression in DPN
two-thirds (62.7%) had at least one chronic musculoskeletal pain condition: osteoarthritis (33.7%) and nociceptive low back pain (26.7%) were the most common; 25.5% had at least one other neuropathic pain condition: carpal tunnel syndrome (13.7%) and low back
Table 1 Demographic and Clinical Characteristics
Characteristic Mean (SD)/n (%)
Age (years) 18--34 6 (2.4%) 35--44 18 (7.1%) 45--54 49 (19.2%) 55--64 67 (26.3%) 651 104 (40.8%) Missing 11 (4.3%) Average age 61.3 (12.8)
Sex Male 114…