Which Multi-dimensional Pain Inventory Subscales Suggest Depression in Veterans with Spinal Cord Injury?
Brian Mutchler, PsyD; Michael Ellwood, PhD; Brenda Scott, PhD; Cathy Williams-Sledge, PsyD; and William Carne, PhD
Abstract
The interplay between depression and pain is noted in clinical settings. Nearly 1 in 5 persons with spinal cord injury (SCI) report pain. The pain-depression relationship was examined in patients with SCI at a Veterans Administration Medical Center by reviewing 156 patient charts. Patients completed a Multi-dimensional Pain Inventory (MPI), Beck Depression Inventory (BDI). A step-wise regression analysis revealed BDI scores were best predicted by a model consisting of Pain Interference, Life Control, and Affective Distress subscales of the MPI. The MPI Interference variable alone significantly predicted BDI scores. High pain interference reports warrant assessment for depression.
Introduction
The interplay between depression and pain is commonly noted in many clinical settings. Theorists have postulated any number of relationships ranging from the actual causation of pain by emotional factors to a more dynamic and subtle relationship between psychic and somatic states. While pain and depression are frequently thought to co-exist in spinal cord injury (SCI), many males with SCI within the VA system do not readily endorse depressive symptoms. Men have been found to be more reluctant to seek help for depression and tend to endorse social withdrawal and somatic concerns compared to women (Galdas, Cheater & Marshall, 2005; Oliver & Toner, 1990). Consequently, depressed mood may be significantly under reported in male veterans with SCI and chronic pain.
In a review of studies, Fishbain, Cutler, Rosomoff, and Rosomoff (1997) concluded that the hypothesis of depression developing as a result of chronic pain was most commonly supported. Weaker support was found for the idea that cognitive attributions associated with pain mediate depression onset. Additionally, there is growing consensus by researchers that depression and chronic pain share some of the same neurotransmitters and pathways, supporting the long-standing clinical impression of a mind-body connection (Gallagher & Cariati, 2002). The emergence of serotonin and norepinephrine as possible mediators of both pain and depression has intrigued both researchers and clinicians alike. The emerging understanding of the roles of neurotransmitters holds promise.
The identification and diagnosis of depression often relies, at least in part, upon the patient’s report of disturbances in sleep and appetite as well as more general reports of anergia, anhedonia and malaise. These subjective complaints often have a somatic focus and may, in fact, be even more specific (e.g., gastrointestinal complaints, headache, backache, neck pain). Thus depression, even in its actual diagnosis, does not always reflect pure psychic phenomena. One study (Simon, VonKorff, Piccinnelli, Fullerton, & Ormel, 1999) of 1,146 patients suggested that somatic issues might actually be the prime (or only) complaint of patients with major depression. This raises the concern that if patients present with narrowly defined somatic complaints the depression may go unrecognized (Kirmayer, Robbins, Dworkind, & Yaffe, 1993).
The outcome of unrecognized depression is often very serious. Numerous medical conditions are adversely effected by untreated depression (Gavard, Lustman, and Clouse, 1993; Frasure-Smith, Lesperance, and Talajic, 1995; Pohjasvaara, Vataja, Leppavouri, Kaste, and Erkinjuntti, 2002). Depression may also increase the perception of pain and decrease psychosocial functioning (Holzberg, Robinson, & Geisser, 1996). A recent study headed by Eugene Carragee (2004) found psychological distress to better predict who would develop back pain than did imaging and diagnostic discography injection. These negative outcomes of unrecognized depression presumptively have major personal, familial, societal and fiscal costs. A biopsychosocial approach is conceptually sounder, because single-modality treatments are rarely effective in pain relief (Cole, 2002).
Co-morbidity of pain and depression has significance for the person with SCI. Mariano (1992) found pain reports ranged widely from 18% to 63%. Methodological problems and inconsistencies between studies may account for much of this variability. Even so, a conservative estimate suggests nearly one in five persons with SCI report significant pain. Mariano suggested a psychosocial evaluation/model as the best intervention for pain intensity.
In general, psychosocial factors have been found to be at least as important, if not more important, than physical injury characteristics in pain intensity. For example, Summers, Rapoff, Varghese, Porter, and Palmer (1991) concluded that anger and negative affect were associated with greater pain perception and complaint. A similar conclusion regarding level of injury and pain was reached by Widerstrom-Noga, Felipe-Cuervo and Yezierski (2001) who looked at the degree of pain interference with sleep, work, exercise, household chores, and other daily activities of SCI patients. Level of injury was not well associated with degree of pain, but level of pain interference with life functioning was highly related to pain levels. Tate, Forchheimer, Maynard and Dijkers’ (1994) findings also support the association between restricted life activities and depression. A 2004 study by Rintala, Hart and Priebe found a continuing relationship between stress and chronic pain in SCI.
Clinicians then would be well advised to determine when psychological variables might be active in the perception of SCI pain. Yet in general “very little has been written regarding chronic pain as a secondary problem in persons with a disability such as spinal cord injury. . . ” (Ehde & Jensen, 2004). The authors further opine, “clinicians working with persons with disabilities would do well to identify those with chronic pain and provide them with opportunities to engage in biopsychosocial treatments. . . ” (p. 257). The literature supports a similar dynamic association between chronic pain and negative psychosocial experiences in persons with SCI, as has been found in the pain literature more globally.
The present study examines the co-occurrence of pain and depression in SCI patients in a large Veterans Administration hospital population. Our goal was to identify the pain characteristics of patients that would signal the need for assessment for depression. Achieving this with greater efficiency might allow for more effective interdisciplinary treatments and prevent declines in health associated with mood disorders.
Method
Participants
The 80-bed Spinal Cord Injury and Disorders unit at the Hunter Holmes McGuire Veteran’s Affairs Medical Center provides life-long care to military veterans living with spinal cord injuries or disorders (SCI/D) in the Mid-Atlantic region. Services include physical rehabilitation, nursing care, caregiver training, health psychology, equipment needs assessment and provision, driver training, wound care, acute medical care, pain management and preventative care. Approximately 1,000 veterans receive their care from the McGuire SCI unit on an ongoing basis.
Veterans with SCI at the Richmond McGuire VAMC are routinely given annual evaluations that include an assessment of the patient’s psychological functioning, involving the administration of both the Beck Depression Inventory (BDI) and the Multidimensional Pain Inventory (MPI). We retrospectively examined the charts of 156 consecutive patients who received annual evaluations from June 2001 through March 2003. Only those charts that contained scorable test instruments were included.
Materials
The Beck Depression Inventory (Beck & Steer, 1987) is a 21-item instrument which takes ten minutes to complete and requires a sixth grade reading level. It purports to measure symptoms and attitudes associated with clinical depression.
The Multidimensional Pain Inventory (Rudy, 1989) measures psychosocial functioning as well as behavioral variables and was designed to be applicable across a wide range of pain types, locations and etiologies. The instrument contains 61 items and has been shown to be sensitive to change in psychosocial functioning. The instrument has been comprehensively studied and validated. In general, the inventory measures three clusters: Family Support and Interaction, Pain Intensity, and Activity levels. Specifically, thirteen scales are scored: Pain Severity, Interference, Life Control, Affective Distress, Support, Punishing Responses, Solicitous Response, Distracting responses, Household Chores, Outdoor Work, Activities Away from Home, Social Activities and General Activity Level.
A brief description of each of the 13 MPI scales follows:
1. Pain Severity––perception of the degree of pain
2. Interference––the patient’s perception as to how pain interferes with their life and their level of current satisfaction
3. Life Control––Perceived control over pain and life events, ability to cope
4. Affective Distress––rating of mood, irritability and tension
5. Support––appraisal of the amount of support received from significant other
6. Punishing Responses––significant other ignores, expresses irritation, anger, or frustration
7. Solicitous Responses––significant other gives pain medications, food, takes over chores and asks how help may be given
8. Distracting Responses––significant other suggests or encourages diversionary hobbies
9. Household Chores––patient helps with meal preparation, household chores
10. Outdoor Work––patient cuts lawn, gardens, works on car
11. Activities Away from Home––patient goes out to eat, takes trips, goes to movies
12. Social Activities––patient visits friends, plays cards
13. General Activity––A composite scale based on # Scales 9–12
Procedure
As part of the annual evaluation standard protocol, each veteran was asked to complete a Multi-dimensional Pain Inventory (MPI) and Beck Depression Inventory (BDI). Institutional Review Board (IRB) approval was obtained for the authors to review the charts, with respect to scores on the MPI and BDI, as well as age and ethnicity and to enter scores into a database.
Data were analyzed using SPSS (SPSS, 1999). Analysis first consisted of descriptive statistics of the patient age, BDI scores and MPI subscale scores. Second, Pearson correlations (using a two tailed test, with significance level set at p < .01 were calculated between MPI subscales, age and BDI scores. Thirdly, stepwise regression analysis was used to predict which variables best predicted BDI scores.
Results
One hundred and fifty-six patient charts were reviewed as part of this study; one patient was female. Of the 156 subjects, 93 were Caucasian, 62 were African American and one person was unidentified as to ethnicity. Some veterans did not elect to complete all of the instruments or (due to other annual evaluation procedures) did not have time to complete all instruments––accounting for the varying numbers cited in Table 1.
As Table 1 indicates, the sample consisted primarily of middle-aged males (mean age = 52.5 years), who when pooled, endorsed only very mild depressive symptomlogy (mean BDI = 10.58) as measured by the BDI. MPI subscale T-scores for affective distress, pain severity, and pain interference were nearly one standard deviation lower than the mean of chronic pain patients upon which the MPI was normed. The MPI subscale T-scores for general activity, life control and social support were essentially equivalent to the MPI pain sample.
MPI scales were inspected with respect to correlation to the Beck Depression Inventory as well as age. Moderate significant correlations between the BDI were found with the MPI Pain Severity Scale (r = .469), Pain Interference (r = .60), Life Control
(r = -.516), and Affective Distress (r = .576). Age was only correlated with General Activity (r = -.404).
A step-wise regression analysis of the data revealed that BDI scores were best predicted by a three variable model consisting of the Pain Interference, Life Control, and Affective Distress subscales of the MPI (adjusted R2 = .463). An ANOVA of the model was significant (F3, 113 = 34.289, p < .01). The Pain Severity, Support, and General Activity subscales of the MPI did not account significantly for variance in the BDI scores. Notably the MPI Interference variable alone significantly predicted BDI scores (F1, 115 = 71.455, p < .01) with an adjusted R2 = .378.
Discussion
The patients with SCI/D in our sample exhibited only very mild depressive symptoms. Likewise, the MPI results suggest that our overall SCI sample does not experience or report pain severity, interference or affective distress at rates as high as the MPI normative sample, which was comprised of chronic pain clinic patients (Rudy, 1989). While reporting less pain and distress, the veteran SCI population has nearly equivalent functioning to the MPI normative sample with respect to general activity, social support and life control. This finding requires more investigation.
While the level of depression in the sample appears relatively low overall, certain aspects of chronic pain are associated with depression. This finding has importance to the front line clinician assisting individuals with SCI because the temporal sequence is, in some respects, irrelevant. The fact is the patient has pain––often unalleviated by analgesic medication––and clinicians need to help the patient manage the pain more effectively. The current study underscores the importance of examining the patient for depression if persistent chronic pain complaints are identified. Conversely, providers of psychosocial interventions for depressed patients should periodically assess for pain complaints.
Key factors may be the perceived interference of pain with life functioning. It appears that individuals with perceived interference in life functions from pain are most at–risk for the development of depressed mood. A focused interpretation of the MPI might prove an efficient method of identifying underlying depression in persons with SCI. Any elevation on Pain Interference, Life Control, or Affective Distress should be considered markers for follow up assessment of underlying depression due to their moderate correlations with the BDI. In our view, a good pain assessment should not limit itself to the quantification of severity (e.g., the standard 1-10 scale of pain severity), but should also include the patient’s perception of how pain interferes in the activities of daily living and the pain’s impact on emotional health.
The present data have significant limitations, particularly with respect to generalization to a wide variety of clinical settings. Due to the mostly male population of a VA hospital, issues relating to female patients with SCI and chronic pain are not illuminated to any degree by this study. Males typically self-report less depression than females. Another factor involves the limitations of the correlational method. Based on our clinical experience of pain complaints being voiced more often than depressive comments, we selected the BDI as the dependent variable. This statistically replicates the interpersonal process commonly observed in the clinical setting. Correlational methods do not, of course, indicate directional causality. Finally, due to a scarcity of fiscal and personnel resources, many available demographic variables were not captured in this study. While we do believe that our sample of convenience likely reflects the general veteran SCI population, this is reasoned speculation––not empirically supported. Subsequent studies will, hopefully, include these data.
References
Beck, A. T. and Steer, R. A. (1987). Manual for the Beck Depression. San Antonio, TX: The Psychological Corporation.
Carragee E. J., Barcohana B., Alamin T., & van den Haak E. (2004). Prospective controlled study of the development of lower back pain in previously asymptomatic subjects undergoing experimental discography. Spine, 29(10), 1112-1117.
Cole, E. B. (2002). Pain management: Classifying, understanding, and treating pain. Hospital Physician, 38(6), 23-30.
Ehde, D. M. & Jensen, M. P. (2004). Feasibility of a cognitive restructuring intervention for treatment of chronic pain in persons with disabilities. Rehabilitation Psychology, 49(3), 254-258.
Fishbain, D., Cutler, R., Rosomoff, H. & Rosomoff, R. (1997). Chronic pain-associated depression: Antecedent or consequence of chronic pain? A review. Clinical Journal of Pain, 13, 116-137,
Frasure-Smith, N., Lesperance, F., Talajic, M. (1995). Depression and 18 month prognosis after myocardial infarction. Circulation, 91, 999-1005.
Galdas P. M., Cheater F. & Marshalll P. (2005). Men and health-seeking behavior: Literature review. Journal of Advanced Nursing, 49(6), 616 623.
Gallagher, R. & Cariati, S. (2002, October 2). The pain-depression conundrum: Bridging the body and mind. Medscape, Retrieved December 30, 2004 from http:www.medscape.com/viewprogarm/2030?src=search.
Gavard, J. A., Lustman, P. J., & Clouse, R. E. (1993). Prevalence of depression in adults with diabetes: An epidemiological evaluation. Diabetes Care, 16, 1167-1178.
Holzberg, A. D. Robinson, M., & Geisser, M. E. (1996). The effects of depression and chronic pain on psychosocial and physical functioning. Clinical Journal of Pain, 12, 118-125.
Kirmayer, L. J., Robbins, J. M., Dworkind, M., & Yaffe, M. J. (1993). Somatization and the recognition of depression and anxiety in primary care. American Journal of Psychiatry, 150, 734-741.
Mariano, A. J. (1992). Chronic pain and spinal cord injury. Clinical Journal of Pain, 8(2), 87-92.
Oliver, J. & Toner, B. B. (1990). The influence of gender role typing on the expression of depressive symptoms. Sex Roles, 22(11-12), 775-790.
Pohjasvaara, T., Vataja, R., Leppavouri, A., Kaste, M. & Erkinjuntti, T. (2002). Cognitive factors and depression as predictors of poor outcome 15 months after stroke. Cerebrovascular Disorder, 14, 228-233.
Rintala, D. H, Hart, K. & Priebe, M. (2004). Predicting consistency of pain over a 10-year period in persons with spinal cord injury. Journal of Rehabilitation Research and Development, 41(1), 75-88.
Rudy, T. (1989). Mutiaxial assessment of pain, multidimensional pain inventory. Computer Program Manual Version 2.1. Pittsburgh, PA: University of Pittsburgh Press.
Simon, G. E., VonKorff, M., Piccinelli, M., Fullerton, C. & Ormel, J. (1999). An international study of the relation between somatic symptoms and depression. New England Journal of Medicine, 341, 1329-1335.
SPSS Inc. (1999). SPSS Base 10.0 for Windows User’s Guide. Chicago: SPSS Inc.
Summers, J. D., Rapoff, M., Varghese, G., Porter, K., & Palmer, R. (1991). Psychosocial factors in spinal cord injury pain. Pain, 47, 183-189.
Tate, D., Forchheimer, M., Maynard, F., & Dijkers, M. (1994). Predicting depression and psychological distress in persons with spinal cord injury based on indicators of handicap. American Journal of Physical Medicine and Rehabilitation, 73(3), 175-183.
Widerstrom-Noga E. G, Felipe-Cuervo, & Yezierski, R. P. (2001).Chronic pain after spinal injury: Interference with sleep and daily activities. Archives of Physical Medicine and Rehabilitation, 82(11), 1571-1577.
Brian Mutchler, PsyD; Michael Ellwood, PhD; Brenda Scott, PhD; Cathy Williams-Sledge, PsyD; and William Carne, PhD, are licensed clinical psychologists at Hunter Holmes McGuire Veterans Affairs Medical Center in Richmond, Virginia. Dr. Carne is also affiliated with the Department of Physical Medicine and Rehabilitation at Virginia Commonwealth University, Richmond, Virginia.