| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||||||||||
Correspondence: Address correspondence to Steven H. Zarit, Department of Human Development and Family Studies, The Pennsylvania State University, Henderson S–211, University Park, PA 16802. E-mail: Z67{at}psu.edu
 |
Abstract |
|---|
 TOP Abstract MethodsResultsDiscussionReferences |
|---|
Key Words: Chronic pain • Very late life • Functional impairment
By virtue of living longer, older adults have an increased risk of developing a wide array of diseases and disorders that have pain as a central component, including osteoporosis, osteoarthritis, diabetic neuropathies, and cardiovascular problems (James, Large, Bushnell, & Wells, 1991; Parmelee, 1994). This risk could be expected to continue to increase among the very old, who are people aged 85 and older. This age group is important, because the number of people in it is expected to increase considerably in the United States and most European countries (Kinsella & Velkoff, 2001). With high rates of chronic illness and disability, the oldest old consume a disproportionate share of health care services (Rosenwaike, 1985; Thorslund & Kundberg, 1994). Little is known, however, about the amount of pain experienced in this age group and its effects on daily experience. In the present study, we examine the frequency and stability of pain over time in a representative sample of the oldest old, qualitative aspects of the pain experience, and the implications of pain for daily functioning.
Epidemiological Studies of Pain
Epidemiological studies indicate that chronic pain is a significant health problem for a large number of older individuals (Brattberg, Thorslund, & Wilkman, 1989; Sternbach, 1986; Von Korff, Dworkin, Le Resche, & Kruger, 1988). Estimates of the prevalence of pain among people aged 65 and older, however, vary considerably, ranging between 35% and 86% (Louis Harris and Associates, 1985; Mobily, Herr, Clark, & Wallace, 1994; Roy & Thomas, 1987; Sorkin, Rudy, Hanlon, Turk, & Steig, 1989; Walker, Akinsanya, Davis, & Marcer, 1990). The most common types of pain involve joint problems secondary to osteoarthritis or rheumatoid arthritis (Louis Harris and Associates, 1985; Marsland, Wood, & Mayo, 1976; Walker et al., 1990). Some studies report that rates of pain increase with age (e.g., Crook, Rideout, & Brown, 1984; James et al., 1991). Other studies, however, have found that the oldest respondents report less pain, or that some types of pain (such as headaches and back pain) decrease with age and other types (such as joint and heart pain) increase (Marsland et al., 1976; Von Korff et al., 1988).
In a population-based Swedish sample, Brattberg, Thorslund, and Wilkman (1989) indicated that the prevalence of chronic pain was 34% for the 18–44 age group, 50% for the 45–64 age group, and 36% for the 65–84 age group. In a more recent study, Brattberg, Parker, and Thorslund (1997) extrapolated from cross-sectional data to posit a curvilinear relationship between age and pain, with the prevalence of pain increasing up to the age of 50, decreasing through age 85, and then increasing once again. Nearly 50% of their sample of oldest old (77–98 years of age) reported multiple locations of pain, and 33% rated the pain as severe. Among those with pain, age was positively related to severity of pain for men and negatively related to total pain for women.
These counterintuitive findings of a decrease in pain from midlife up to age 85 have several possible explanations. It may be that there are cohort differences, with older generations less likely to report pain. Alternatively, older people may have either a higher threshold for experiencing pain or may accept their pain as part of everyday life in old age.
Correlates of Pain
There is relatively little information concerning the correlates of pain in the oldest old. In younger age groups, researchers have associated chronic pain with a variety of negative outcomes, including decreases in physical functioning (Lipton & Stewart, 1993), impairment of activities of daily living (Lavsky-Shulan et al., 1985), reduction in social and leisure activities (Deyo & Tsui-Wu, 1987), mortality (Moss, Lawton, & Glickman, 1991), and suicide (Penttinen, 1995; Stenager, Stenager, & Jensen, 1994). Depression in particular has been identified as a frequent concomitant of chronic pain (Fishbain, Cutler, Rosomoff, & Rosomoff, 1997; Romano & Turner, 1985; Rudy, Kems, & Turk, 1988; Turk, Okifuji, & Scharff, 1995). Incidence rates of depression in pain patients range from 10% to 100% (Roy, Thomas, & Matas, 1984). Some researchers have found small though significant relations between pain and depression among nursing home and congregate apartment residents (Cohen-Mansfield & Marx, 1993; Parmelee, Katz, & Lawton, 1991). Rudy and colleagues (1988), however, contend that the presence of pain is an insufficient condition for the subsequent development of depression. They found that the relationship between depression and pain was mediated by a decline in the perception of control and mastery, and perceived reductions in instrumental activities of daily living (see also, Okifuji, Turk, & Sherman, 2000).
As at any age, pain among older people reflects pathological processes. Many physicians and elders, however, dismiss legitimate pain complaints, attributing them to a natural, nonpathological part of the aging process. Furthermore, studies indicate that physicians typically underestimate pain among nursing home residents, especially those who are cognitively impaired or nonverbal (Sengsten & King, 1993). The failure to identify pain can have dire consequences for the quality of life of pain sufferers. Misdiagnosed or untreated pain can lead to excess disability, functional impairment, and depression (Carrol & Bowsher, 1993). Given the multiple losses endured by the oldest old, it becomes critical to accurately assess and understand the nature of pain in very late life.
Our purpose in the current investigation was to examine the prevalence and incidence of pain in a short-term longitudinal study of the oldest old; to delineate the nature of the pain experience within the lives of the participants; and to investigate the effects of pain on other areas of health and functioning. The study had four objectives. Our first objective was to determine the prevalence and descriptions of pain in a population of adults over the age of 85. Our second objective was to assess changes in pain perceptions over a 2-year period. Our third objective was to examine the correlates of pain, including demographic factors that might account for differences in reported pain, as well as the consequences of pain for health and functioning. Our fourth objective was to examine factors associated with changes in pain perception over the 2-year period.
|
Methods |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
Using census data, we drew a stratified random sample. Each stratum consisted of individuals born in the following birth years: 1897, 1899, 1901, and 1903. Of people contacted for the study, the initial participation rate was 86% (see Johansson & Zarit, 1995, or Zarit, Johansson, & Malmberg, 1995, for complete descriptions of the sample). The sample includes individuals who reside in independent community housing, as well as institutional settings such as nursing homes and assisted living facilities. For the present investigation, participants were initially interviewed about their pain experiences when they were aged 86 to 92 (baseline interview), and then again 2 years later (follow-up interview).
In this article, we focus on the people who completed the pain questions at both times of measurement. Of the 213 participants interviewed at baseline, 190 (89%) completed the pain assessment. The remaining 23 participants did not participate in this part of the study, mainly because they had severe cognitive or sensory impairment. Of the 190 people who gave pain assessments at baseline, 98 (52%) completed pain assessments at follow-up, 71 (37%) were deceased, 6 (3%) refused to be reinterviewed, and 15 (8%) were interviewed but could not complete the pain questions, usually as a result of severe cognitive impairment. Thus, the main reason for attrition between baseline and follow-up was mortality.
Given the high attrition between baseline and follow-up, it is possible that survivors may have differed in sociodemographic characteristics likely to influence pain, such as gender or education. It is also possible that people with higher pain at baseline were less likely to complete the follow-up assessment. To test for these attrition effects, we divided the sample into three groups: (a) people who completed pain assessments at baseline and follow-up; (b) people who died before follow-up; and (c) people with pain assessment only at baseline or those who refused to be interviewed at follow-up. These latter two groups were combined because of the small number of cases in each. As shown in Table 1, people who completed pain assessments at both waves were more likely to live in independent housing. A higher proportion of women were in the group that completed both assessments compared with the group that died, but the highest proportion of women was found among those who did not complete the second assessment. Initial measures of reported pain were not associated with either mortality or nonparticipation at follow-up.
|
Demographic Information
We obtained demographic data, including the individual's education, marital status, gender, age, and type of residence (independent or institutional).
Assessment of Pain
We assessed pain in a two-step process. We used an initial screening question to determine the presence and frequency of pain. For participants who reported any pain, interviewers asked a series of additional questions about the characteristics of the pain and their experience of it.
In the initial screening question, interviewers asked participants to describe the frequency with which they experienced pain in their lives. Responses were made on a 5-point Likert-type scale, with the following choices: never, only sometimes, on a weekly basis, on a daily basis, or continuously. A higher score indicates more frequent pain.
To determine the reliability of this approach, we compared responses on the screening question to diagnoses of illnesses associated with chronic pain (e.g., osteoarthritis or cancer) and with current use of analgesics. (In Sweden, analgesics are usually obtained with a prescription, because the cost would then be covered by health insurance.) Of the 190 persons for whom pain data were available at baseline, only 6 (3%) were found with either a diagnosis or medication typically associated with pain who did not respond affirmatively to this screening question. Because pain is a subjective experience and the medications in question could have had other purposes, we did not reclassify any participants; instead we relied on the initial screening question to divide the sample into people who are experiencing pain and those who are not.
People who reported any pain in response to the initial screening question were asked to complete a pain "map." We used the pain map to assess the location of pain, the total number of painful sites, and the qualitative aspects of the pain experience. We adapted the pain map in this study from the McGill Pain Questionnaire (Melzack, 1987). The map consists of a drawing that represents a dorsal and frontal view of a human body. Interviewers directed participants to mark those areas of the body in which they experienced pain. Interviewers also asked participants to give a verbal descriptor for the pain they were experiencing, choosing from a list of common types of pain: aching, burning, numbness, piercing, muscle cramp, or shooting pain. A symbol accompanied the verbal descriptor of each type of pain and was placed on the site or sites of the body where the pain was experienced.
High test–retest reliability (.85) for pain maps has previously been reported (Margolis, Chibnall, & Tait, 1988). Pain maps and drawings have exhibited concurrent validity with the Pain Rating Index, Number of Words Chosen, and Present Pain Index subscales of the McGill Pain Questionnaire (Margolis, Tait, & Krause, 1986). Margolis and colleagues (1986) also established interrater reliability, which ranged from.92 to.98. In the present study, two raters evaluated the pain maps subsequent to the interview. Interrater agreement was.95 for presence of pain in a particular part of the body and.97 for type of pain.
Participants who reported any pain were asked about the duration of pain, specifically how long they had been experiencing pain. This question was asked about pain in general, rather than pain in any specific site. Other questions included the degree to which pain influenced participants' way of life, the degree of influence they have over their pain, and the perceived helpfulness of the treatments in alleviating their pain.
Health
We assessed health by using ratings of subjective health and information about participants' medical conditions and the medications they used. We measured global subjective health with four questions that rated participants' current health, the severity of their health conditions, their health now compared with that of 2 years ago, and their health compared with that of other people the same age. We scored responses for participants' current health on a 7-point Likert scale with choices anchored by very poor on one end and excellent on the other. We scored the other items on 3-point scales. A factor analysis indicated that these items formed a single factor, with loadings between.43 and.89. Because the items had different response choices, we computed a factor score and used it in the analysis. Higher scores indicate better perceptions of global subjective health. Evidence from validation studies shows that subjective health correlates with objective measures of global health (Rodin & McAvay, 1992).
To assess medical conditions, interviewers asked participants to report current illnesses. The total number of illnesses is reported. The interviewers also asked about sleep disturbance, including difficulty falling asleep and waking in the middle of the night. To determine medication use, nurse-interviewers asked participants to show them any prescription medications they were taking. The number and type of prescription medications were recorded.
Depression
To examine the relation of pain and depression in this sample, we assessed the current level of depressive symptomatology with a short version of the Center for Epidemiological Studies–Depression scale (CES-D). Participants reported the frequency with which they felt depressed or downhearted and other similar items derived to tap depressive symptoms. The CES-D short form has high internal reliability with alphas in the.85 to.90 range and correlates highly with the full version of the CES-D (Kohout, Berkman, Evans, & Cornoni-Huntley, 1993). Cronbach's alpha for this sample was.87 at baseline and.82 at follow-up; 10% of the sample fell into the range of clinically significant depression at each assessment.
Global Cognitive Ability
We used the Mini-Mental State Exam (MMSE; Folstein, Folstein, & McHugh, 1975) to assess global cognitive ability. The MMSE is a brief screening instrument used to determine cognitive functioning in several domains. Scores range from 0 to 30; higher scores are indicative of better functioning. A score of 23 or below is generally considered an appropriate cutoff for dementia. At baseline, 23% of the sample fell below this level, indicating possible dementia. This figure rose to 43% at follow-up. The standardized Cronbach's alpha coefficient for the MMSE in this sample was.90 at baseline and.88 at follow-up.
Disabilities in Daily Living
We measured activities of daily living (ADLs) for three domains: instrumental ADLs (IADLs), physical ADLs (PADLs), and mobility. We took items from widely used scales to form a comprehensive assessment in each of these domains (Katz, Ford, Moskowitz, Jackson, & Jaffe, 1963; Lawton, 1971). The standardized Cronbach's alpha coefficient for each of the subscales was greater than.95 at baseline and follow-up. Higher scores indicate better functioning.
Analyses
We performed the analyses in three steps. The first objective of the study was descriptive in nature; thus prevalence, frequency, and the qualitative aspects of pain were described for both times of measurement. Second, to examine the longitudinal course of pain perception, we looked at changes in the prevalence, frequency, and qualitatitve aspects of pain. We also examined change by categorizing participants into four groups: those who had (a) no pain at both times of measurement, (b) pain at follow-up only (incident cases), (c) pain at baseline only (recovered cases), and (d) pain at both times of measurement. Third, we looked at correlates of pain at baseline and follow-up, and we included whether demographic characteristics accounted for differences in reported pain and if pain was associated with other areas of health and functioning. Fourth, we examined differences associated with changes in pain over time by using a series of one-way analyses of variance. Because people with cognitive impairment might be less reliable reporters of pain, in particular, underreporting the occurrence and severity of pain experiences, we also reported findings for people with cognitive impairment (an MMSE score of 23 or lower) and without impairment (an MMSE score of 24 or higher).
|
Results |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
2 = 2.28(1), p =.14.
|
|
Changes in reported pain from baseline to follow-up are shown in Table 3. Of the sample, 51% had no pain at both times, whereas 24% reported pain at both assessments. Among the remaining sample, 16% were incident cases of pain at follow-up, and 9% no longer reported any pain. Frequency of pain from baseline to follow-up was moderately correlated (Spearman's 
=.50; p <.001). Cognitive status was not significantly associated with whether people reported incident pain at follow-up (12% of cognitively impaired, compared with 17% of nonimpaired who reported incident pain) or no longer reported pain at follow-up (12% of cognitively impaired, compared with 7% of nonimpaired).
People with incident pain at follow-up reported a frequency of pain of 2.69 (SD = 1.01), which falls between weekly and daily on the rating scale, and a mean of 2.19 painful sites (SD = 2.51). The frequency was similar to that reported at follow-up by people with pain at both times (
= 2.87, SD = 0.81). People with incident pain, however, reported fewer pain sites than people with pain at both measurements: 2.19, compared with 4.43, t(35) = 2.02, p <.05. People with pain at baseline but not follow-up had previously reported a mean of 4.56 sites (SD = 4.88).
Duration of pain at baseline ranged from less than 1 month to 80 years (this was reported by one person; the next highest duration was 20 years). Mean duration of pain was 75.61 months and the median was 29 months. At follow-up, the duration of pain ranged from less than 1 month to 360 months (30 years). The mean duration was 89.73 months and the median was 60 months. Using a period of 3 months or more as the criterion for chronic pain (Merskey, 1986), we could classify 98% of the sample reporting pain at both times of assessment as suffering from chronic pain.
An examination of the stability of reports of duration of pain from baseline to follow-up indicated some problems. Approximately one half (54%) of respondents with pain at both times of measurement reported a duration of pain follow-up that was consistent with the amount of time between interviews; that is, they now estimated the duration of their pain as 1.5 to 2.5 years longer than at baseline, which is consistent with the amount of time that passed between interviews. Eighteen percent reported a duration of pain at follow-up that was identical to the amount reported at baseline or of a shorter duration, despite the passage of 2 years. Twenty-eight percent of the sample reported a greater duration of pain at follow-up than was accounted for by the 2-year interval. These expanded duration estimates ranged from 4 to 20 years. The group reporting an increase in duration consistent with the passage of time tended to have pain of relatively shorter duration (all but two had reported pain initially of less than 5 years). The biggest outlier reported a duration of 80 years at baseline and 30 years at follow-up. Consistency in these reports was unrelated to cognitive impairment, with similar proportions of cognitively impaired among the consistent and inconsistent reporters. Thus, these estimates of the onset of pain have to be viewed with some caution. In part, this problem reflects the chronicity of pain in the sample, and the consequent difficulty in pinpointing the exact time of onset.
Table 4 shows the types of pain reported. Aching pain was the most frequent type of pain experienced at baseline (59%) and follow-up (33%) and affected the most areas of the body.
|
Correlates of Pain at Baseline and Follow-Up
We examined whether demographic measures at baseline accounted for differences in the frequency of pain (continuous to never) and number of painful sites at baseline and follow-up (Table 5). Age and education were not associated with either pain measure. Among other demographic variables (not shown in Table 5), women reported a similar frequency of pain as men at baseline ( = 4.03 compared with 4.22) and follow-up ( = 3.92 compared with 3.81). Men and women also reported similar number of painful areas at baseline ( =1.40 compared with 1.48), but at follow-up women reported a higher amount: = 1.63 compared with 0.70, t(96) = 2.12, p <.05. Marital status was not related to pain measures at baseline but was found to be related to the number of pain sites at follow-up: = 0.64 for married people, 1.63 for unmarried, t(96) = 2.38, p <.05. Type of housing (independent vs. institutional) was not associated with the pain measures.
|
Factors Associated With Changes in Reported Pain
Finally, we considered the factors at baseline that might be associated with changes in reported pain. We evaluated change by using the four categories reported earlier: no pain at both times, pain at baseline only (recovered), pain at follow-up only (incident), and pain at both times of measurement. Given the small number of cases involved in some cells (see Table 3), these analyses must be viewed as preliminary. Among demographic variables, only age was associated with changes in pain: F(3, 94) = 5.78, p <.001. The recovered group was somewhat older than other groups ( = 91.33 years at baseline), whereas the incident group did not differ from the no pain or continuous pain groups ( = 89.13 for the incident group, compared with = 88.16 for no pain and 88.61 for continuous pain). Turning to the relation of changes in pain to the health variables, we found that there were significant differences among the groups in the number of medications at baseline, F(3, 94) = 5.10, p <.01, and subjective health at baseline, F(3, 89) = 5.12, p <.01. People with continuous pain or pain only at baseline were taking more medications than other groups. People with no pain at both times reported better subjective health. The four groups did not differ significantly in their MMSE scores at baseline or follow-up. Trends were apparent, however, for people in the no pain at both times or recovered groups to have lower MMSE scores at follow-up. People who recovered (no pain at follow-up) had MMSE scores of 24.33 (SD = 5.05) at baseline and 21.44 (SD = 3.71) at follow-up. People with no pain at both times had scores of 24.85 (SD = 5.05) at baseline and 20.86 (SD = 8.40) at follow-up. By comparison, people with pain at both times had MMSE scores at baseline of 25.78 (SD = 3.59) and at follow-up of 25.27 (SD = 3.49) and the incident pain group had MMSE scores of 23.13 (SD = 7.29) at baseline and 21.07 (SD = 8.40) at follow-up. Thus, there is some indication that people not reporting pain at follow-up were declining more in MMSE scores compared with people reporting pain at both times, though incident cases also showed some cognitive decline.
We also examined changes in pain reports for their relation to ADL functioning and mobility. Mean scores of the four groups did not differ significantly either at baseline or follow-up on PADLs or on IADLs. PADLs were F(3, 93) = 1.23, p =.30 at baseline and F(3, 93) = 0.56, p =.64 at follow-up. IADLS were F(3, 93) = 0.73, p =.54 at baseline and F(3, 93) = 0.72, p =.54 at follow-up. There were, however, significant differences in mobility at baseline, F(3, 92) = 3.35, p <.05, with the people having no pain having the highest mobility ( = 18.98) and those who subsequently recovered having the lowest ( = 13.78). Changes in pain were not associated with depression scores at either time of measurement.
|
Discussion |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
Prevalence of reported pain in this sample of the oldest old was somewhat lower than that typically found in studies of people aged 65 and older (Mobily et al., 1994; Roy & Thomas, 1987). The prevalence rates were also lower than the one previous study of the oldest old (Brattberg et al., 1996), but the types of pain reported were similar. A major difference between our findings and other studies is that we used a two-step process in assessing pain. Participants were initially asked a screening question about whether they were currently troubled by aches or pains. They were then asked the complete pain inventory only if they reported any pain on that initial question. In contrast, most prior studies have focused primarily on pain and have asked all participants to respond to the full inventory of possible types of pain and pain sites. As an example, all participants in the study by Brattberg and colleagues (1996) were read a list of body parts and were asked to indicate the presence or absence of pain for each location. That method may elicit reports of mild or occasional pain that a respondent might otherwise not note. Indeed, the prevalence rate of pain in our study corresponds to the rate for severe pain reported by Brattberg et al. (1996). As noted, we examined the reliability of pain reports in our study by looking at the relation of reported pain to use of pain medications and diagnoses of illnesses such as cancer and osteoarthritis that are generally characterized by pain. The reports of pain by participants in our study were consistent with their use of pain medications and diagnoses of illnesses generally characterized by pain. As a result, we believe the method for identifying pain to be reliable and to reflect people who experience pain on more than an occasional and mild basis. Even the higher rate reported by Brattberg et al. (1996), however, is less than that typically found in samples of people aged 65 and older, suggesting that pain may be somewhat less prevalent in very late life.
We also found a modest increase in reports of pain over time. It may be, as Brattberg and colleagues (1997) suggest, that reports of pain decrease from age 50 to 85, and then begin increasing. This trend may indicate cohort differences in expectations about pain, thresholds in experiencing pain in later life, or both. It is also possible that the rates of reported pain found in the oldest old may, in part, reflect selective survivorship; that is, that people who experience severe pain earlier in life do not live to advanced old age.
Although participants reported that their pain had a considerable effect on their life, the magnitude of the effect of pain on other areas of functioning was generally modest. The strongest associations were with other health measures, such as subjective health, number of medications, and sleep disturbance. In terms of everyday functioning, pain had significant but modest correlations with depression and mobility, whereas associations with PADLs and IADLs were not significant. Thus, pain contributed to but was not an overwhelming determinant of functioning in everyday life. It is likely, then, that pain in very late life is only one among several factors that can lead to poor functioning.
Demographic variables had few significant associations with pain. The lack of relation to age and education is due, at least in part, to the narrow range on both those variables. More surprising was the lack of a gender difference, which is a common finding in reports of pain at earlier ages (Andersson, Ejlertsson, Leden, & Rosenberg, 1993; von Korff et al., 1988). Using cross-sectional data, Brattberg and colleagues (1996) reported that pain increased with age among men and decreased for women. These findings might indicate changes in the experience of pain with advancing age, or, again, selective survival.
Dementia is very common in this age group and cannot be ruled out as a source of unreliability in reporting pain symptoms. The proportion of people with significant cognitive impairment (an MMSE score of 23 or lower) nearly doubled in this sample from baseline to follow-up. Whereas people with the most severe cognitive impairment did not participate in the pain interview, it is possible that some people with mild or moderate impairment might have underreported pain experiences. In the analyses, we found that pain reports were similar in people with and without cognitive impairment. Higher frequency of pain at baseline, however, was associated with better cognitive functioning at follow-up. Simmons and Schnelle (2001) reported that people with mild to moderate cognitive impairment are able to report their experiences in a reliable way. Nonetheless, we cannot rule out the possibility that some people with cognitive impairment underreported the occurrence or severity of pain. We need a more thorough examination of this issue to clarify the reliability of self-reports of people with cognitive impairment.
A related issue is that the group of people who "recovered" from pain was somewhat older and more frail than the group of people continuing to report pain; that is, they had lower MMSE and mobility scores. Although the number of cases was too small to explicate these differences fully, it may be that at least some of these individuals were having difficulty reporting their pain.
Another factor that may account for the moderate prevalence of pain is that the oldest old may minimize or ignore some of their experiences of pain, particularly when it is minor and transitory. As suggested by Johnson and Barer (1996), the oldest old may be protected by an "aura of survivorship" by which their long life conveys special status and allows them to minimize the threats to their daily life and well-being. In a longitudinal panel, Ågren (1998) reported a similar protective quality in how the oldest old adapt to limitations and disabilities. It is also conceivable that these individuals have learned to live with their pain or feel less compelled to complain than younger adults or elders of other ethnic backgrounds.
The main limitation of the present study is the small sample size. Future studies should also include additional pain measures. Missing in the present battery were verbal descriptor scales designed to measure the intensity of the pain experienced. A modification of the complete McGill Pain Questionnaire (Margolis et al., 1988) or the Short Form McGill Pain Questionnaire (Margolis et al., 1986) used in conjunction with the pain measures in this study would lead to a comprehensive assessment of pain in late life. That type of approach could also clarify if the two-step assessment of pain used in the present study accounted for the relatively low prevalence of pain compared with other studies, or if pain is less prevalent in very late life.
|
Footnotes |
|---|
1 Department of Human Development and Family Studies, The Pennsylvania State University, University Park.
2 Rehabilitation Research and Development Center of Excellence on Geriatric Rehabilitation, Atlanta VAMC, Atlanta, GA.
3 Institute of Gerontology, University of Jönköping, Sweden.
Decision Editor: Laurence G. Branch, PhD
Received for publication January 21, 2002. Accepted for publication September 21, 2002.
|
References |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  J. F. Karp, J. W. Shega, N. E. Morone, and D. K. Weiner Advances in understanding the mechanisms and management of persistent pain in older adults Br. J. Anaesth., July 1, 2008; 101(1): 111 - 120. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
