Introduction

Musculoskeletal pain is highly prevalent in the elderly, and individuals with musculoskeletal pain frequently report pain in multiple sites [1, 2]. Studies show that about 25–43% of community-dwelling adults aged 65 years and older report musculoskeletal pain in two or more sites, representing a significant burden is this growing population [1, 3]. Compared to older persons without pain or with a single site of pain, those with multisite pain are more at risk for disability and falls [3,4,5,6,7]. In addition, multisite pain is associated with low balance confidence [8], and reduced physical performance [1, 9].

Geriatric syndromes are common chronic conditions in older adults that are not defined by discrete disease processes [10]. They are often multifactorial, sharing similar risk factors and pathways [10]. Falls, urinary incontinence, functional disability, delirium, and frailty are examples of geriatric syndromes, typically caused by the collective effects of multiple organ damage in an aging person [10, 11]. Although each geriatric syndrome is unique, they share similar risk factors and overlapping pathways [10, 12]. Recent evidence suggests that multisite pain may derive from multiple systemic factors, such as chronic inflammation and central pain sensitization related to brain plasticity, rather than localized disease processes [13, 14]. Multisite pain, similar to established geriatric syndromes, typically is not defined by a discrete disease process yet is associated with substantial suffering and health risks in older individuals.

In this study, we hypothesized that the risk factors for multisite pain would bear many similarities to other established geriatric conditions such as falls, activities of daily living (ADL) disability, frailty, and urinary incontinence. We propose that multisite pain should be added to the list of geriatric syndromes requiring clinical assessment and management to prevent functional loss. This study aims to determine the prevalence and characteristics of multisite pain in relation with other geriatric syndromes and to evaluate whether multisite pain may represent a distinct geriatric syndrome.

Methods

The MOBILIZE Boston Study (MBS), “Maintenance of Balance, Independent Living, Intellect and Zest in the Elderly”, is a population-based cohort study of older adults that enrolled 749 participants aged 70 and older. At the start of the study, participants lived in the Boston area and nearby suburbs within a 5-mile radius of the Institute for Aging Research at the Hebrew Rehabilitation Center (HRC). Age-eligible people from the city and town lists were randomly selected and recruited door-to-door.

Initial MBS eligibility was based on the following: aged 70 years and older, ability to understand and communicate in English, plan to be in the area for at least 2 years, and ability to walk 20 feet without personal assistance [15]. Individuals were excluded from the study if they were diagnosed with a terminal disease, or if they had moderate to severe cognitive impairment using the Mini-Mental Status Examination (MMSE score, < 18) [16]. The participants provided written informed consent at the beginning of the baseline assessment, which was conducted in parts, an in-home interview followed within 2 weeks by a clinic assessment. Further details on the recruitment and methods were published previously [15]. The Institutional Review Board of the HRC approved the study protocol and consent procedures.

Chronic pain

A 13-item joint pain questionnaire was used to assess chronic musculoskeletal pain in hands and wrists, shoulders, back, chest, hips, knees, and feet [17]. Chronic pain at each site was determined if the pain was present in the previous month and present for at least 3 months in the previous year, regardless of laterality [7]. Anginal chest pain, assessed using the Rose questionnaire, was excluded from the musculoskeletal pain classification [18]. Chronic pain was classified as follows: (1) multisite pain (≥ 2 sites), (2) pain in a single site, and (3) no pain [7].

Falls

History of falls was based on the participant’s report of having fallen to the ground or on the stairs one or more times in the previous year [19].

ADL disability

Self-reported disability is referred to as report of any difficulty in performing one or more activities of daily living (ADL) including bathing, dressing, transferring, using the toilet, and eating [20].

Frailty

The frailty index was based on Fried’s frailty phenotype derived using data from the Cardiovascular Health Study (CHS) [21]. An algorithm was developed in the MBS to classify participants according to the five frailty criteria: weight loss, exhaustion, low physical activity, slow walking speed, and weakness. Self-reported unintentional weight loss was based on the interview question, “In the last year, have you lost more than 10 pounds unintentionally, that is, not due to dieting or exercise?” Weakness was defined as the highest quintile of the time to perform five chair-stands without using one’s arms, adjusted for BMI and gender. This test is a component of the Short Physical Performance Battery (SPPB) [22]. Slow gait was based on performance in the highest quintile of time for a usual-paced 4-m walk, also a component of the SPPB. Exhaustion was determined from response to the Center for Epidemiologic Studies Depression Scale (CESD) item, “Over the past week or so, did you feel like you could not get going?” [23]. Participants who reported this symptom for three or more days in the past week were classified as having exhaustion. Low physical activity was measured as the lowest activity score quintile from the Physical Activity Scale for the Elderly [24]. Frailty was classified based on the presence of three or more of the five criteria [21].

Urinary incontinence

Urinary incontinence was based on an affirmative response to a single item of reported “trouble with any involuntary leakage of urine or unable to hold urine during the past year”.

Demographic and health characteristics

Demographic characteristics included age, gender, and race. Body mass index (weight in kilograms divided by height in meters squared) was calculated from measured height and weight. Standard cut points for overweight (25–29.9) and obesity (≥ 30) were applied. The Mini-Mental Status Examination was used to assess general cognitive status ranging 18–30, using the cut point of 24 to indicate normal function [16]. Heart disease was determined based on self-report of heart attack, congestive heart failure, pacemaker, or cardiac arrhythmia, and also the presence of angina, assessed using the Rose questionnaire [7]. The presence of diabetes was based on an algorithm that included self-reported diabetes mellitus, use of anti-diabetic medications, and laboratory testing for an elevated random glucose level (≥ 200 mg/dl) and hemoglobin A1c ≥ 7% [7]. Osteoarthritis of the hand and knee were assessed by a research nurse trained in the American College of Rheumatology clinical criteria [25, 26]. Other musculoskeletal diseases were self-reported, including rheumatoid arthritis, spinal stenosis, and disc disease. Multiple comorbidity refers to the co-existence of two or more chronic conditions based on self-report of 12 major chronic conditions [27].

Depression was measured using a validated modification of the 20-item Center for Epidemiologic Studies Depression Scale (CESD-R) [23, 28]. A letter chart, the Good-Lite Chart model 600A, was used at a distance of 10-feet to measure distant vision [7]. Anxiety symptoms were measured using the reliable and well-validated 7-item anxiety subscale of the Hospital Anxiety and Depression Scale (HADS) [29].

Short Physical Performance Battery (SPPB)

The SPPB, used to measure mobility performance, includes 3 tests, the 4-m usual-paced walk, standing balance, and repeated chair stands [22]. Gait speed was measured as the shortest time of two trials of the timed 4-m walk. Standing balance was assessed using three timed, 10-s stands requiring the participant to stand with their feet side-by-side, semi-tandem (standing with the side of the heel of one foot touching the side of the big toe of the other foot), tandem (standing with a heel to toe). The repeated chair stands test measured the ability and time required to stand up and down from a chair five times as fast as possible with arms folded across the chest. The SPPB score ranged from 0 to 12, calculated by summing the three test scores based on established cut points [22]. Scores less than ten (< 10) indicated poor mobility function [22].

Medications

The interviewer during the home visit recorded all prescription and over-the-counter drugs taken in the previous 2 weeks based on review of the medication bottles and participant’s self-report. Use of psychiatric drugs was based on any use of anxiolytics, anti-depressants, sedative/hypnotics, or antipsychotic medications in the previous 2 weeks.

Statistical analysis

Descriptive statistics were used to examine the following seven groups of variables: demographic measures, health measures, and multisite pain, ADL disability, urinary incontinence, falling, and frailty. Chi-square tests were used to compare the prevalence of risk factors according to multisite pain and other geriatric syndromes.

The first stage in the analysis involved assessing the bivariate association between potential risk factors, specifically age, gender, race, lower SPPB, heart disease, diabetes mellitus, poor vision, lower MMSE, anxiety, depression, psychiatric drugs, and multiple comorbidities, in relation to the multilevel pain outcome. Subsequently, a series of multinomial logistic regression models were performed to determine multivariable-adjusted associations between demographic and health characteristics and categories of pain. All analyses were performed using STATA version 14 (StataCorp LLC, College Station, TX, USA).

Results

The average age of participants was 78 ± 5 years (range 70–97 years), 63% were female, and 77.5% were White. Participants enrolled in the MBS were largely representative of the overall demographic profile of older adults living in the Boston area based on comparisons with data from the 2000 US Census [15].

Of the 749 participants enrolled in the study, 300 (40%) reported chronic multisite pain (Table 1). Almost half of the MBS population (49%) met criteria for frailty. Urinary incontinence was reported by 41% of participants, while 37% reported having fallen in the past year and 23% reported ADL disability. All five geriatric conditions were associated with poor mobility performance and depression (p ≤ 0.05). Neither age nor race was associated with pain. More women than men had multisite pain and urinary incontinence. Although prevalence of multiple comorbidity was high in general in this population, it was highest among those with any of three geriatric conditions, multisite pain, ADL disability, and frailty.

Table 1 Demographic and health characteristics according to pain status and geriatric syndromes, MOBILIZE Boston Study, 2005–2007 (n = 749)
Full size table

In an examination of the overlapping quality of geriatric syndromes, we found that multisite pain was prevalent in about half of participants with urinary incontinence (48%), fall history (48%), and frailty (49%), and in 3 out of 5 (61%) people with ADL disability (Fig. 1). Of note, multisite pain was reported by 26% of those who did not have another geriatric syndrome.

Fig. 1
figure 1

Pain prevalence among older adults according to the presence of other geriatric syndromes (n = 749)

Full size image

Musculoskeletal pain in any one site was commonly part of a multisite pain condition. Among those with chronic pain in any selected musculoskeletal site, between 75 and 90% reported pain in one or more other pain sites (Fig. 2). In addition, the prevalence of chronic musculoskeletal conditions was highest among those with multisite pain, with 70% of older adults with multisite pain having at least one musculoskeletal disease (Fig. 3). However, only about one-third (35%) of those with multisite pain had conditions that would be expected to cause multisite musculoskeletal pain, such as rheumatoid arthritis or osteoarthritis at more than 1 site. Alternatively, 30% with multisite pain did not report or meet criteria for any major musculoskeletal disease. Among those with a single site of pain or no pain, 48% and 22%, respectively, had one or more musculoskeletal diseases (Fig. 3).

Fig. 2
figure 2

Percent of older adults with multisite pain according to specific pain sites (n = 749). Participants reporting more than one chronic pain site, determined by self-report of pain in the previous month, present for at least 3 months in the previous year

Full size image
Fig. 3
figure 3

Prevalence of articular diseases according to pain category, (n = 749). Mutually exclusive groupings according to the following hierarchy (descending order): osteoarthritis and spinal disease or disc disease, osteoarthritis > 1 site, rheumatoid arthritis, disc disease/spinal disease, osteoarthritis of knee, osteoarthritis of hand. Brackets indicate conditions affecting multiple musculoskeletal sites

Full size image

After adjusting for several chronic conditions, demographic characteristics were not associated with either single site or multisite pain (Table 2). We found a number of health measures and chronic conditions were independently associated with multisite but not single site pain, including mobility limitations (low SPPB score), multiple comorbidity, overweight or obesity, anxiety and depression. We did not find that any demographic or health characteristics were independently associated with single site pain.

Table 2 Demographic and health risk factors for multisite and single site pain, 749 older adults, MOBILIZE Boston Study
Full size table

Discussion

Geriatric syndromes are a constellation of chronic conditions that exact a heavy toll in the lives of older adults. Multisite pain is highly prevalent, affecting 40% of older adults in our study, similar to the proportions living with other recognized geriatric syndromes. As hypothesized, we found that risk factors for multisite pain were quite similar to factors associated with other recognized geriatric conditions. Demographic and health characteristics associated with multisite pain and other geriatric syndromes included gender (female), depression, number of comorbid conditions, and limited mobility performance. In contrast, we did not find that presence of single site pain was associated with any identified risk factors for multisite pain. Hence, in contrary to multisite pain, single site pain may not represent part of a geriatric syndrome.

To our knowledge, this is the first population-based study to demonstrate the prevalence and characteristics of multisite pain in the context of other geriatric syndromes. Our prevalence estimate for multisite pain in the older population is similar to reports from other US and UK populations [2, 3]. If an older person reports pain in any selected site, they are very likely to have multisite pain. Similar to other studies [2, 3], we observed that consistently around 80–90% who report pain in one joint or back area actually had multisite pain.

When Tinetti and colleagues explored risk factors for three geriatric syndromes, falls, incontinence and ADL dependence, they observed that the syndromes shared a number of risk factors including anxiety, limited vision and hearing, and poor strength and mobility (chair stands) [11]. In addition, body mass index was related to urinary incontinence and functional dependence while comorbidity (≥ 2 chronic conditions) was associated with all of these geriatric syndromes. Using a similar approach, we found that a number of these shared risk factors were independently associated with multisite pain, including anxiety, depression, poor physical performance, multiple comorbidity, and obesity. The findings of our risk factor analysis support our hypothesis about the similarities between multisite pain and other geriatric syndromes.

Olde Rikkert and colleagues identified norms for classifying geriatric syndromes [30] and again, our study provides evidence that multisite pain meets this proposed classification. Our findings support that multisite pain is similar to other established geriatric syndromes such as falls, urinary incontinence, and delirium, as follows: (1) multisite pain is a highly prevalent condition in the older population, (2) pain in multiple locations is related to different causes or impairments in other systems, (3) a substantial overlap exists between multisite pain and risk factors of other geriatric syndromes, and (4) many individuals with multisite pain also have other geriatric syndromes. Therefore, multisite pain in older adults appears to be a geriatric multifactorial condition associated with several underlying risk factors, multisystem decline, and age-related chronic diseases.

Multisite pain in older adults is often considered as an expected, almost normal part of aging by both patients and the health care providers. This attitude can contribute to insufficient evaluation and management of multisite pain in older patients. Since multisite pain often is not linked to a single disease or specific condition, identifying multisite pain as a geriatric syndrome sheds light on this condition as multifactorial, burdensome for elderly adults, and requiring more standardized approaches to care. Phelan and colleagues proposed that recognition of geriatric syndromes could lead to more appropriate alternative strategies to the usual disease-based approach to care [31]. They proposed a “risk factor assessment and reduction approach” to management because the multifactorial symptoms of geriatric syndromes derived from chronic incurable problems [31]. Therefore, addressing shared risk factors is the key to reducing the burden and disablement associated with co-occurring geriatric syndromes. With multisite pain, consistent with geriatric guidelines, both pharmacologic and nonpharmacological approaches are needed as a long-term approach to effective management [32]. In addition, recognition of multisite pain as a geriatric syndrome could lead to further research to develop more thorough pain assessments and more effective strategies to manage multisite pain as an identifiable condition in older adults. Hopefully such recognition would achieve the ultimate goal of reducing the overall burden of pain and improving function among older adults who live with this condition.

In 2011, the Institute of Medicine’s landmark report, Relieving Pain in America, drew attention to the problem of chronic undertreatment of pain and the burden this places on individuals, families, and society [32]. Health care providers also need to have a better understanding of atypical manifestations and consequences of multisite pain in older patients. Improved pain management requires understanding the many approaches available for treating pain in older patients and recognizing the pharmacokinetic and pharmacodynamic changes that occur with aging [33, 34].

There are a few limitations to be considered in this study. First, the MOBILIZE Boston study population did not include the most disabled adults living in the community (i.e., those who could not walk without assistance) and those with moderate or severe cognitive impairments. Self-report data have limitations though, as a subjective symptom, pain assessment always requires self-report. The pain assessment was limited to chronic pain in musculoskeletal joint areas and the back which are among the most common sites of chronic pain leading to disability in older adults [1, 4]. Since pain in the abdomen or head was not included, we cannot confirm the role these sites might play in a global pain syndrome. In addition, there are other measures of frailty such as the Clinical Frailty Scale based on a more comprehensive assessment which was not available in our data [35]. Finally, the cross-sectional nature of the study limits conclusions about temporal relationships between the multisite pain and other syndromes. The prospective relationship between pain and disability has been shown previously [4] but future studies should explore associations with other syndromes longitudinally. Strengths of the study include the population-based approach, extensive assessment of pain and other geriatric syndromes, extensive health information, and generalizability of our findings to similar populations of older adults.

Conclusion

Multisite pain is a common problem associated with substantial burden and risk in the older population. With the rapid expansion of the older population in coming decades, the burden of multisite pain will become a more urgent concern. Although prior studies have explored risk factors for multisite pain, these new findings reveal that multisite pain, often unexplained, bears many similarities to established geriatric syndromes. Thus, multisite pain warrants further consideration as a unique geriatric syndrome, requiring new approaches to evaluation and management of this condition.