Relationships between Cognitive Screening Composite Scores and Pain Intensity and Pain Disability in Adults with/at Risk for Knee Osteoarthritis

Abstract

Objectives:

Chronic pain, cognitive deficits, and pain-related disability are inter-related. The prevalence of chronic pain and undiagnosed cognitive difficulties in middle age and older adults is increasing. Of the cognitive systems, executive function and episodic memory are most relevant to chronic pain. We examined the hypothesis that cognitive screening composite scores for executive function and memory would negatively associate with pain intensity and pain disability in a group of middle aged and older adults with knee pain, with or at risk for osteoarthritis.

Methods:

A total of 120 adults (44 M/76 W), average age of 59 years participated in the study. Demographic, health history, clinical pain, and cognitive measures were completed. Relationships between pain intensity, pain disability, and the Montreal Cognitive Assessment (MoCA) total and composite scores were examined with relevant covariates in the model.

Results:

MoCA raw scores ranged from 13 to 30 with a mean score of 23.9. Pain intensity was negatively associated with overall MoCA total, and executive function and memory composite scores. Pain disability over the previous six months was negatively associated with executive function while pain disability over the past 48-hours was not associated with executive function.

Conclusions:

The current study demonstrates associations between pain metrics and cognitive domain scores within a common cognitive screening tool.

Introduction

There is an increasing need for cognitive screening within clinical pain treatment settings. Researchers show a reciprocal relationship between chronic pain and cognitive dysfunction [1, 2] such that individuals with cognitive deficits and chronic pain report more pain-related disability relative to individuals without cognitive deficits [3, 4]. The prevalence of chronic musculoskeletal pain is also increasing in middle aged and older adults [5, 6], as are rates of undiagnosed cognitive difficulties in community-dwelling older individuals [7]. By 2050, it is estimated middle age to older adults will represent more than 1.6 billion of the world population [8] and our healthcare system will face even greater numbers of individuals meeting mild to moderate neurocognitive disorder diagnoses [9]. For these reasons, understanding and assessing relationships between cognition, pain intensity, and pain disability in community-dwelling middle-age and older adults may assist with the development of prevention interventions, improved functioning, and reduced long-term costs.

Of the primary cognitive systems, executive function and episodic memory may be most relevant to understanding associations between cognition, pain intensity, and pain disability. Executive function is broadly defined as the cognitive processes to complete critical everyday tasks like paying attention, staying focused and regulating emotions [10]. Episodic memory is defined as the ability to store information about an event including the time and place [11]. Executive function and episodic memory are prominently implicated in chronic pain research [12, 13] and in many forms of mild to moderate neurocognitive disorders. Executive function and chronic pain overlap in frontostriatal pathways, which evaluate relevance of pain stimuli and are key region for modulation of higher order cognitive function [14]. Cognition and chronic pain are also both associated with mesocorticolimbic pathways, which show reduced functional connectivity in chronic pain [15–17], aid in shifting attentional sets, suppressing old behavioral sets [10], and prioritizing information for memory consolidation [18]. Importantly, the frontostriatal and mesocorticolimbic pathways change with normal and abnormal aging [19].

To date, research demonstrating associations between pain intensity and pain disability with these cognitive domains has relied upon neuropsychological research protocols [20]. Although some researchers have examined cognitive screening measures for executive and memory domains on outcomes such as dementia classification [21, 22] or delirium prediction [23], examination specific to chronic pain and functioning would be highly relevant. Findings may have potential clinical value for appreciating pain intensity and pain disability patterns in older adults, as recent studies show that individuals experiencing executive dysfunction, but not memory dysfunction, are more likely to experience pain disability [24].

The current investigation examined the relationships between a cognitive screening test total score, and executive function and memory composite scores in a group of middle aged and older adults with knee pain with or at risk for osteoarthritis. In prior studies, individuals with reduced executive function have shown increased interference from chronic pain independent of pain intensity [24] as well as relationships between neuronal pathways of pain modulation, memory, and executive function [14, 15–17]. Therefore, we hypothesized that pain intensity would negatively associate with both executive function and memory composites, while pain disability would associate with an executive function composite alone.

Methods

Participants and Procedures

This study is a secondary data analysis of participants from a larger, multisite (University of Florida, University of Alabama at Birmingham) study examining ethnic/race group differences in pain and function among individuals with or at risk for knee osteoarthritis (OA; Understanding Pain and Limitations in Osteoarthritic Disease [UPLOAD-2]) [25]. Individuals aged 45 and older who self-identified as non-Hispanic Black or non-Hispanic White were recruited from the community via posted fliers, radio and print media announcements, orthopedic clinic recruitment, and word-of-mouth referral. Data for this study were collected between 2015 and 2017. The University of Florida Institution Review Board and the University of Alabama at Birmingham Institution Review Board reviewed and approved the study. All participants provided written informed consent and were compensated for their involvement. Only participants from the University of Florida site who completed a cognitive assessment and completed all pain measures and reported knee pain consistent with or at risk for knee osteoarthritis were included in this analysis. Data were collected during the baseline health assessment visit. Participants completed health history and pain questionnaires and a brief cognitive assessment. Questionnaires were reviewed with participants to reduce missing information and collection errors. Study staff were trained in administering the Montreal Cognitive Assessment (MoCA) for consistency in conducting the assessment. This study follows the STROBE statement for reporting cross-sectional studies [26].

Measures

Baseline Characteristics

Participants provided information including age, sex, self-reported ethnicity/race, highest education level attained, income level, and total number of current comorbidities including high blood pressure, heart disease, cancer, diabetes, asthma/breathing problems, kidney disease, thyroid problem, stroke, seizure, chronic pain, neurological disorder, depression, other mental health condition or other health problems. Total number of pain sites was collected based on bilateral areas including hands, arms, shoulders, neck, head/face, chest, stomach, upper back, lower back, knees, legs (other than knees), and/or feet/ankles (0-24 sites).

Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC)

The Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC; [27, 28] assesses knee OA symptoms over the past 48 hours using a 4-point Likert scale. The WOMAC is comprised of three subscales: pain (0-20 score), physical function (0-68 score) and stiffness (0-8 score). Higher scores indicate worse pain, impairment of physical function, or stiffness. A WOMAC total score is derived by adding the scores of the three subscales. For the purposes of this study, only pain and physical function subscale scores were included in the analyses, as these were the constructs of interest. The WOMAC has been reported to have Cronbach’s alpha coefficient of 0.86 – 0.89 among patients with knee and hip OA [28].

Graded Chronic Pain Scale (GCPS)

The Graded Chronic Pain Scale (GCPS) [29] assesses knee pain over the past 6 months with seven items based on a 10-point Likert scale. The GCPS includes the GCPS characteristic pain intensity and GCPS disability scores amongst other scores. Characteristic pain intensity is derived by averaging current, worst, and average knee pain during the past 6 months. The disability score is the degree to which chronic pain interferes with daily activities, social/recreational activities, and work activities. The two scores are computed by averaging their items and multiplying by 10 to generate index scores for characteristic pain intensity (0-100 score) and disability (0-100 score), with higher scores indicating greater symptomatology. To reduce the number of categories for the subsequent ordinal regression analysis, GCPS disability scores were categorized according to previously published cutoffs to create a disability score from 0 to 3 (0-29=0, 30-49=1; 50-69=2, 70-100=3; [29]). GCPS pain intensity remained at 0 to 100 due to no identification of a published categorical conversion.

Montreal Cognitive Assessment (MoCA)

The Montreal Cognitive Assessment (MoCA) is a widely used cognitive screening measure [27]. It has been utilized in many clinical populations including individuals with chronic pain [30]. The MoCA is composed of seven domains totaling 30 points: Visuospatial/Executive, Naming, Attention, Language, Abstraction, Delayed Recall, and Orientation. A total MoCA score of 26 or above is considered normal while under 26 is suggestive of cognitive impairment, though a population-based found that middle-aged individuals in a large population-based sample scored 23-24 on average [31]. We developed two composites by combining cognitive domains for the analyses: 1) all domains requiring cognitive flexibility, including Visuospatial/Executive, Attention, Language, and Abstraction (heretofore referred to as the “cognitive flexibility composite”; 0-16 score) and 2) both domains reflecting semantic and verbal memory, i.e., the Naming and Delay Recall domains (heretofore referred to as the “memory composite”, 0-8 score).

Statistical Analyses

Descriptive statistics are presented as mean/median depending on normative distribution or percentage of the sample for baseline characteristics and outcome measures. Ordinal regression analyses were used to assess the influence of chronic pain on specific cognitive composites. Output from ordinal regression analysis is expressed as an odds ratio, with ratios greater than 1 indicating higher levels of an independent variable increase the odds of the dependent variable being at a higher value, and odds lower than 1 indicating higher levels of an independent variable decrease the odds of the dependent variable being at a higher value.

All analyses controlled for age, sex, level of education (coded as 12 years or less, 13-16 years, and greater than 16 years), and ethnicity/race (Non-Hispanic Black or non-Hispanic White). A total of three pairs of regression analyses for each of the two pain measures (i.e., the WOMAC and GCPS) assessed the relationships between pain and the MoCA total and cognitive composites. The first of these analyses assessed the relationship between pain intensity (i.e., WOMAC pain and GCPS characteristic pain intensity) and the MoCA total and cognitive composites. The second assessed the relationship between pain disability (i.e., WOMAC physical function and GCPS disability score) and the MoCA total and cognitive composites. A third set of analyses repeated the analyses from the second set, while additionally controlling for pain intensity for each of the two pain measures. An alpha <0.05 was considered statistically significant for all testing procedures. All analyses were completed in SPSS v25 (©IBM SPSS Statistics).

Results

Descriptive and Clinical Pain Characteristics

The final sample consisted of 120 participants (44 men and 76 women) with average age of 58.7 ± 8.20 years. Approximately half of the participants identified as Non-Hispanic White (51.7%). Sixty-two participants (51.7%) completed a high school degree or lower, 38 (31.7%) completed an undergraduate degree, and 20 (16.7%) completed a graduate or professional degree.

Participant median pain scores were 53.33 (IQR=36.67-73.33) on the GCPS characteristic pain intensity, 43.33 (IQR=17.50-76.25) on the GCPS disability prior to score conversion and 1 (IQR=0-3) after score conversion, 7 (IQR=4-11) on the WOMAC pain, and 22 (IQR=12-35) on the WOMAC physical function.

Overall MoCA Score and Demographic Characteristics

MoCA scores ranged from 13 to 30 with a mean score of 23.90 ± 3.76. See Table 1. Participant MoCA median score was 24 (IQR=22-27). Higher educational attainment was associated with a higher probability of higher total MoCA score (OR=2.48, 95% CI=1.57-3.94, p<.01). Age and sex were not associated with total MoCA score.

Table 1:

Baseline Characteristics

Variable	Total sample (N=120)
Sociodemographics
Sex, N (%)
Men	44 (36.7%)
Women	76 (63.3%)
Ethnicity/Race, N (%)
NHB	58 (48.3%)
NHW	62 (51.7%)
Age, M ± SD	58.7 ± 8.20
Waist/Hip Ratio, M ± SD	0.90 ± 0.09
No. Comorbidities (0-14), N (%)
0	19 (15.8%)
1-2	78 (65.0%)
3+	23 (19.2%)
Education, N (%)
High School or Less	62 (51.7)
Undergraduate	38 (31.7)
Graduate/Professional	20 (16.7)
Income, N (%)
$0-29,999	67 (55.8%)
$30,000-79,999	33 (27.5%)
$80,000+	16 (13.3%)
Not Reported	4 (3.3%)
Pain
GCPS CPI, M ± SD	54.27 ± 23.98
GCPS Disability, M ± SD	46.97 ± 31.98
WOMAC Pain, M ± SD	7.50 ± 4.50
WOMAC Physical Function, M ± SD	23.91 ± 15.36
Cognition
Cognitive Flexibility Composite, M ± SD	11.68 ± 2.80
Memory Composite, M ± SD	6.27 ± 1.51
MoCA Total, M ± SD	23.90 ± 3.76

NHB=Non-Hispanic Black; NHW=Non-Hispanic White; GCPS CPI= Graded Chronic Pain Scale Characteristic Pain Intensity; GCPS Disability=Graded Chronic Pain Scale Disability; WOMAC Pain= Western Ontario and McMaster Universities Osteoarthritis Index Pain Subscale; WOMAC Disability= Western Ontario and McMaster Universities Osteoarthritis Index Disability Subscale; Cognitive Flexibility Composite = comprised of the sum of Visuospatial/Executive, Attention, Language, and Abstraction from the MoCA; Memory Composite= comprised of the sum of Naming and Delayed Recall domains from the MoCA; MoCA= Montreal Cognitive Assessment

MoCA and Pain Measures

Pain Intensity

WOMAC Pain.

After controlling for age, sex, educational attainment, and ethnicity/race, higher WOMAC pain score was associated with higher odds of a lower MoCA total score (OR=0.87, 95% CI=0.80-0.94, p=.01). When MoCA cognitive composite scores were examined, higher WOMAC pain score was associated with a higher probability of a lower score on both the MoCA cognitive flexibility composite (OR=0.90, 95% CI=0.83-0.98, p<.01) and the MoCA memory composite (OR=0.88, 95% CI=0.81-0.96, p<.01).

GCPS Characteristic Pain Intensity.

After controlling for age, sex, educational attainment, and ethnicity/race, higher GCPS characteristic pain intensity was associated with a higher probability of lower MoCA total score (OR=0.98, 95% CI=0.96-1.00, p=.01). When MoCA cognitive composite scores were examined, higher GCPS characteristic pain intensity was associated with a higher probability of lower scores in the MoCA cognitive flexibility composite (OR=0.98, 95% CI=0.97-1.00, p=.04), but was not significantly associated with the memory function composite (OR=0.99, 95% CI=0.98-1.00, p=.14).

Pain Disability

WOMAC Physical Function.

After controlling for age, sex, educational attainment, and ethnicity/race, higher WOMAC physical function score was associated with higher odds of a lower MoCA total score (OR=0.97, 95% CI=0.94-0.99, p<.01). When MoCA cognitive composite scores were examined, higher WOMAC physical function score was associated with higher odds of lower memory composite (OR=0.98, 95% CI=0.95-1.00, p=.04), and trended towards an association with MoCA cognitive flexibility composite score (OR=0.98, 95% CI=0.95-1.00, p=.06). After controlling for WOMAC pain score, higher WOMAC physical function was not associated with MoCA total score (OR=1.01, 95% CI=0.97-1.06, p=.56), MoCA cognitive flexibility composite (OR=1.01, 95% CI=0.96-1.06, p=.45) or the MoCA memory composite (OR=1.03, 95% CI=0.98-1.07, p=.29).

GCPS Disability.

After controlling for age, sex, educational attainment, and ethnicity/race, GCPS disability was significantly associated with higher probability of lower MoCA total score (OR=0.73, 95% CI=0.55-0.97, p=.03). This association did not remain after controlling for GCPS pain intensity (OR=0.84, 95% CI=0.59-1.19, p=.30). When MoCA cognitive composite scores were examined, higher GCPS disability was significantly associated with lower cognitive flexibility score (OR=0.66, 95% CI=0.50-0.88, p<.01). This association remained significant after controlling for GCPS characteristic pain intensity (OR=0.68, 95% CI=0.48-0.98, p=.04). GCPS disability was not significantly associated with the MoCA memory composite score (OR=0.99, 95% CI=0.74-1.31, p=.92). A summary of findings is presented in Table 2.

Table 2:

Odds Ratios between MoCA Total, MoCA Composites, and Pain Measures

Measures	MoCA Total	Cognitive Flexibility Composite	Memory Composite
Pain
WOMAC	0.87; 0.80-0.94	0.90; 0.83-0.98	0.88; 0.81-0.96
GCPS	0.98; 0.96-<1.00	0.98; 0.97-<1.00	0.99; 0.98-1.00
Disability
WOMAC	0.97; 0.94-0.99	0.98; 0.95-1.00	0.98; 0.95-<1.00
GCPS	0.73; 0.55-0.97	0.66; 0.50-0.88	0.99; 0.74-1.31,
Disability c/v Pain
WOMAC	1.01; 0.97-1.06	1.01; 0.96-1.06	1.03; 0.98-1.07
GCPS	0.84; 0.59-1.19	0.68; 0.48-0.98	1.15; 0.80-1.65

Numbers display OR; 95% CI. Gray shading indicates a significant association, p< 0.05, c/v indicates covariate in the model.

MoCA= Montreal Cognitive Assessment, GCPS= Graded Chronic Pain Scale, WOMAC= Western Ontario and McMaster Universities Osteoarthritis Index

Discussion

The goal of the study was to examine if cognitive composites derived from a common cognitive screening measure, the MoCA, would differentially associate with pain intensity and pain disability in middle age to older adults with knee pain that have, or are at risk for, osteoarthritis. Consistent with our hypotheses, pain intensity over the past 48 hours negatively associated with total MoCA score and the cognitive flexibility and memory composites, while pain disability over the previous six months negatively associated with cognitive flexibility alone even after controlling for pain intensity. Findings have implications for clinical practice as they suggest that individuals with chronic pain who report pain disability are also at greater risk for cognitive flexibility limitations.

The differential associations between the WOMAC and GCPS may reflect differences in the time frame of the two questionnaires. The WOMAC assesses pain and physical function within the past 48 hours, while the GCPS measure collected in the study involves both a current and over the previous six months assessment [27, 29]. The difference in time frame may explain why the WOMAC pain intensity was associated with both executive function and memory; it is a measure of current/recent pain and may therefore be a better measure of attentional resources devoted to pain at the time of the cognitive screening administration. By contrast, the GCPS characteristic chronic pain measure, although including current pain, emphasizes pain over a longer period of time, and may therefore more accurately reflect maladaptive plasticity affecting cognitive function [3]. Therefore, clinicians may benefit from including measures reflecting both recent and longer-term time frames when assessing chronic pain intensity and disability.

Pain processing and cognitive function compete for attentional resources. Detection and modulation of pain is prioritized due to its high biological salience, leaving less attentional resources for cognitive processing. Distraction due to pain also associates negatively with cognitive functioning [3, 32]. Loss of attentional resources affects cognitive flexibility by reducing the ability to attend to multiple stimuli and suppressing automatic responses. It also affects the ability to process and consolidate new information into memory, ultimately affecting episodic memory. This may explain why pain intensity associates with both cognitive flexibility and memory function. Regarding functional difficulties, lowered attentional resources for cognitive flexibility makes organizing and executing daily activities more difficult for individuals with chronic pain. This may explain the association between GCPS pain disability score and cognitive flexibility. The GCPS disability score includes questions focusing on interference of pain with everyday life, work, and social activities. By contrast, the WOMAC’s functional subscale focuses on difficulty with physical movements. This might partially explain the lack of association between WOMAC Physical Function and the MoCA executive composite.

In addition to reduced attentional resources, chronic pain may negatively affect cognition over time through maladaptive plasticity. A complex network of brain regions are involved in pain modulation including the prefrontal cortex, insular cortex, hippocampus, amygdala, periaqueductal grey, and anterior cingulate cortex [3]. Many of these regions, most notably the prefrontal cortex, hippocampus, and anterior cingulate cortex, are also associated with cognitive flexibility and memory. Over time, a bombardment of persistent pain signals could alter the structure of these frontal and temporal networks, ultimately causing cognitive dysfunction. In support of this theory, research has demonstrated reduced gray matter density in frontal and temporal cortices in individuals with knee osteoarthritis [33] and a negative association between OA pain duration and cortical thickness in frontal and temporal regions [34]. Importantly, this process appears to be reversible; several studies have found increases in cortical thickness in frontal regions following treatment for chronic pain [35–37], though cognition was not assessed in these studies. One limitation in the interpretation of MRI findings is that differences may reflect fluctuations in cerebral blood flow [38].

In addition to highlighting differences between measures of pain intensity and pain disability in the context of cognitive domains, evaluating cognitive functioning may help with the identification of clinical interventions to improve pain and cognition outcomes. A randomized control trial found improvements in executive function following computerized cognitive training [39]. Although this intervention did not reduce interference from chronic pain, improving executive function might contribute to improved functional independence [40]. A second study reported improved pain outcomes in response to an 8 week acceptance and commitment therapy intervention for veterans with chronic pain. Importantly, those individuals with relatively lower cognitive functioning, spedcifically, processing speed and executive function, indicated a greater benefit. Thus, lower cognitive functioning would not be an exclusion factor for similar forms of chronic pain treatment [41]. Consideration of cognitive functioning could be beneficial in the treatment of chronic pain. Additionally, the MoCA composites proposed in our study provide a simple and accessible method for further screening for cognitive dysfunction within specific cognitive domains. This is particularly relevant given the brevity, low cost, and lack of training required to administer a MoCA when compared to more traditional methods of assessing executive function. However, further validation will be required to confirm that the composites described in this study are comparable to other executive function measures.

Our study is notable for several strengths. First, the MoCA cognitive composite scores provide an additional method for evaluating executive function and memory domains using a widely administered screening measure. Assessing these cognitive domains may improve the identification of at risk individuals who may benefit from cognitive and/or psychological interventions to help manage chronic pain. Second, we included two well-recognized and validated measures of pain intensity and pain disability to provide a thorough characterization of pain experiences. Third, the diverse representation of the sample in the study increases the generalizability of findings.

There are limitations to acknowledge as well. First, the MoCA executive function and memory composites should be compared to traditional measures. Second, the delayed recall measure on the MoCA reflects spontaneous recall. The measure does not differentiate between deficits in recall (i.e., the ability to remember information without cuing) and recognition (i.e., the ability to recognize previously presented information [42]). Free recall relies on executive function processes as it requires efficient organization of information for recall, while recognition does not [11]. The MoCA has varying degrees of recognition as part of its administration; examinees are given a semantic cue if they cannot recall a word (i.e., “type of flower” for “daisy”), and a 3-word multiple choice option if they cannot recall with a semantic cue. Future studies should consider utilizing these additional scores to better characterize type of memory weakness. Third, our sample consisted of individuals with knee pain with or at risk for osteoarthritis and did not consider other pain conditions. Future research would benefit from 1) considering the contributions of sleep quality and fatigue on observed relationships [43, 44]; 2) completing prospective analyses to better appreciate the direction of relationships between cognitive functioning, pain intensity, and pain disability; and 3) assess cognitive functioning in other chronic pain conditions.

The current study demonstrates the value of considering cognitive domains, even within cognitive screening metrics, in individuals with chronic pain. With a replication of findings, specific cognitive domains, e.g., MoCA composite scores, may serve as a beneficial tool in clinical practice for informing the treatment for people living with chronic pain.