Overlap of pain‐related and general measures of disability among adults with chronic pain

Chronic pain is known to be correlated with disability. We aimed to determine the overlap between a general self‐reported measure of disability and a measure of disability due to pain problems among adults with chronic pain.


I N T RODUC T ION
Chronic pain is defined as pain experienced on most days or every day in the previous three months. 13][4] While research and treatment techniques continue to advance, [5][6][7] most chronic pain conditions remain inadequately understood and managed. 8,9The management of chronic pain typically focuses on restoring functional status and quality of life, 9 as patients with chronic pain are more likely than people without chronic pain to experience a wide range of functional limitations. 2,10For example, in a recent analysis of U.S. population-based data, nearly half of respondents with chronic pain endorsed work limitations due to a health problem, compared with only 15% of people reporting no chronic pain. 2 Although the prevalence of disability and work limitation is significantly higher among people with chronic pain, existing data are limited in identifying to what extent this is specifically due to pain problems. 2,10istorically, population-based studies have analyzed the association of chronic pain with general measures of disability but could not determine whether disability or functional status limitations were caused by pain problems. 11,12A few longitudinal studies have collected population-based estimates of pain interference with regular activities, but these studies were not representative of the entire adult population, being limited to either young adults or older adults. 13,14][17] Beginning in 2019, the National Health Interview Survey (NHIS) has collected population-based data on both general and pain-specific measures of disability among U.S. adults. 2These data permit analyzing the overlap between general and pain-specific measures of functional limitations in a nationally representative sample, including adults not currently receiving treatment for chronic pain conditions.Among adults with chronic pain, we sought to determine how many respondents endorsing a general measure of disability also reported functional limitations specifically due to pain.Our secondary aim was to analyze which demographic and clinical characteristics were associated with reporting (a) both general and pain-specific limitations; (b) only general, but no pain-specific limitations; and (c) neither general, nor pain-specific limitations among adults with chronic pain.

M ET HOD S
The 2019 NHIS was conducted by the National Center for Health Statistics at the Centers for Disease Control and Prevention and was an interview survey sampling people living in households in the United States. 18Past iterations of the NHIS have been used to assess the likelihood of disability in the chronic pain population; 4,19 estimate the prevalence of chronic pain; 2,3,20 and characterize the disability, health status, and healthcare use profile of Americans with chronic pain. 14In each household, one randomly selected "sample adult" (age 18 and older) completed a detailed questionnaire on pain presence, severity, and interference with daily activities.We included "sample adult" respondents who endorsed experiencing pain "most days" or "every day" over the past 3 months. 2dults without chronic pain and respondents with missing data on study variables were excluded from this study.
The primary outcome was determined based on the presence of general activity limitations and painspecific activity limitations. 2,10Pain-specific limitation was defined as reporting that chronic pain limited ones' activities on "most days" or "every day" over the past 3 months.General activity limitation was defined as reporting being limited in the kind or amount of work the respondent could do because of a physical, mental, or emotional problem.Respondents were classified according to whether they reported both general and painspecific limitations; only general, but no pain-specific limitations; only pain-specific, but no general limitations; and neither general, nor pain-specific limitations among adults with chronic pain.
The psychometric properties of these disability measures have not been extensively examined.The question about general activity limitations is based on a longrunning series of questions included in the NHIS about health-related limitations on the kind or amount of work respondents could do.According to one analysis, 88% of Americans receiving disability-based supplemental income report being unable to work at all, or endorse limitations to the amount or kind of work they can do. 21The question related to pain-specific limitations was added to the NHIS as part of the 2019 survey redesign, based on the work of the NIH Interagency Pain Research Coordinating Committee (IPRCC) in defining appropriate measures for capturing the prevalence and burden of chronic pain in the United States. 22dditional measures related to chronic pain included pain severity (evaluated based on the last time the respondent experienced pain, and classified as mild/moderate ["a little" or "somewhere in between a little and a lot"] vs. severe ["a lot"]).Pain locations were classified as reporting "a lot" or "somewhere between a little and a lot" of pain in the back; hands, arms, or shoulders; hips, knees, or feet; abdomen (including pelvic and genital pain); or teeth or jaw not (mutually exclusive).The presence of headache or migraine was assessed using the same definition.
Respondent demographic characteristics included age, sex, race/ethnicity, and place of birth (born in the United States or U.S. territory, vs. foreign-born).The NHIS questionnaire used the terms "sex", "male," and "female"; no other response categories for the sex question were collected or recorded; and no separate questions were asked about gender identity.Other respondent characteristics queried for analysis included type of health insurance (any private coverage; public coverage only; other coverage; or no coverage); relationship status (married; living with a partner; or neither married nor living with a partner), number of children in the family, body mass index (BMI) category (normal or underweight; overweight; or obese, based on self-reported height and weight); and current smoking, lifetime diagnosis of diabetes, and lifetime diagnosis of hypertension.
Data were summarized using weighted means or proportions, as appropriate, and compared across categories of general and pain-related limitation using Wald tests.We used multinomial logistic regression models to analyze how covariates were associated with the likelihood of experiencing both general and pain-related limitation, only general limitations, only pain-related limitations, or no limitations.Multivariable models included all available covariates.Multiple imputation was not used to complete missing data on covariates, due to the low fraction of missing information.All analyses used NHIS weights, and standard errors were adjusted to account for the NHIS complex survey design. 18Stata/ SE 16.2 (College Station, TX: StataCorp, LP) was used to conduct all analyses, and p < 0.05 was considered statistically significant.
Primary data collection for the NHIS was approved by the National Center for Health Statistics Research Ethics Review Board.Deidentified data from the NHIS are made available to the public, and analysis of these public, deidentified data was not considered human subjects research by our local Institutional Review Board.

R E SU LT S
The 2019 NHIS included 31,997 adult respondents, of whom 7184 reported chronic pain.After excluding 310 respondents with missing data on study variables, we retained an analytic sample of 6874 respondents (weighted mean age: 55 years; weighted percent female: 54%).Based on this sample, we estimated that in the population of U.S. adults with chronic pain, 58% had some kind of disability, including 27% with both general and pain-related disability (95% confidence interval [CI]: 26%, 29%), 22% with only general, but not pain-related disability (95% CI: 20%, 23%), and 9% reporting only pain-related, but not general disability (95% CI: 8%, 10%).In other words, among people with both chronic pain and any disability, 62% were experiencing disability specifically related to their pain problem.
Pain severity and location, and management approaches are summarized for each group in Table 1.As expected, severe pain was most common among respondents who reported both general and pain-related disability (61%), followed by respondents who reported only pain-related, but not general disability (48%).Respondents with both general and pain-related disability had the highest likelihood of experiencing pain at each body site covered in the survey, especially hips, knees or feet (75%), and back pain (72%).
Respondent demographic and clinical characteristics are compared by the type of disability reported in Table 2. Compared with other groups, respondents reporting both general and pain-related disability were especially likely to have public insurance coverage and to be single.This group also had the highest rates of obesity, smoking, diabetes, and hypertension among all study groups.Respondents reporting only painrelated, but not general disability tended to be younger, more likely to be women, more likely to have private insurance coverage, and more likely to be married and living with children, compared with other groups of people with disabilities in the study.Respondents reporting only pain-related disability also had lower rates of obesity, smoking, diabetes, and hypertension than respondents reporting both pain-related and general disability.
Further analyses estimated independent associations between each patient characteristic and the likelihood of reporting general and pain-related disability.The main comparison of interest was between patients reporting only pain-related disability but no general disability, and those reporting both pain-related and general disability (Table 3).Severe pain and pain at additional sites were associated with lower likelihood of reporting only pain-related disability.Considering demographic and socioeconomic characteristics, women were more likely than men to report only painrelated disability (relative risk ratio [RRR]: 1.5; 95% confidence interval [CI]: 1.2, 2.0; p = 0.001), whereas respondents with nonprivate insurance coverage were more likely to report both general and pain-related disability.Smoking and presence of comorbidities (diabetes and hypertension) were associated with increased likelihood of reporting both pain-related and general disability, rather than reporting pain-related disability alone.
Considering other pairwise comparisons among study groups, patients with more severe pain and T A B L E 1 Weighted proportions (with 95% confidence intervals) of chronic pain severity and location, according to presence of general and pain-related disability.pain affecting additional body sites were more likely to report both general and pain-related disability than reporting general but no pain-related disability (Table A1).Lastly, Table A2 assesses factors associated with reporting both types of disability as compared to neither.More severe pain and pain at additional sites were associated with increased likelihood of reporting both types of disability, rather than neither; as were older age, nonpublic insurance coverage, smoking, diabetes, and hypertension.All other possible pairwise comparisons of outcome categories in the multivariable multinomial logistic regression analysis are summarized in Tables A3-A5.

DI SC US SION
Chronic pain is associated with increased prevalence of disability. 2,20However, it is unclear how much of this increased disability prevalence is specifically related to pain problems, and therefore might be resolved with improved pain management.By cross-referencing measures of general disability and disability specifically related to pain conditions, we found that 62% of people with chronic pain who disclose a disability are experiencing activity limitations related to their pain problem.Therefore, improving chronic pain management could help close the gap in the prevalence of disability between people with and without chronic pain.Moreover, 9% of all adults with chronic pain reported disability due to their pain problem, but did not endorse a general measure of disability commonly used in population health research.Therefore, assessing disability without specifically measuring pain interference with usual activities has likely underestimated the burden of disability among people with chronic pain.Several demographic and clinical factors were associated with reporting only pain-related disability, and not general disability.Most importantly, women were especially likely to report pain-related but not general disability, suggesting that previous research might have underestimated the contribution of chronic pain to disability among women.4][25][26] Considering the higher prevalence of chronic pain among women, comprehensive assessment of painrelated activity limitations can help address the higher burden of pain-related disability in this population.Additionally, respondents with chronic pain in our study who reported only pain-related disability were less likely to smoke or have comorbidities than those who also reported general disability, suggesting that the impact of pain on daily activities in the former group may be underestimated because of the absence of other comorbidities associated with functional limitations.Previous analyses have identified that lifestyle risk factors such as smoking and hypertension are linked to greater pain severity and functional impairment. 27,28Since chronic pain is known to be more prevalent among adults with lifestyle risk factors, pain-related disability could be missed or underestimated among adults who present without these risk factors.
Our study highlights the importance of understanding complex patient perspectives on functional limitations when treating patients with chronic pain.Reducing activity limitations and helping people get back to work are common treatment goals in rehabilitation. 29,30In some studies, measures of pain interference were specifically targeted as outcomes for improvement with rehabilitation treatment. 31However, other evaluations of rehabilitation for patients with chronic pain only focused on general measures such as return to work or use of sick leave. 32Our results emphasize that from both clinical and research perspectives, tracking reduction in pain interference with usual activities is an important component of assessing the efficacy of rehabilitation for patients with chronic pain, including multidomain interventions recently described in the literature. 33onclusions from our study are limited by several characteristics of the data and analytic approach, including the use of cross-sectional data, reliance on self-reported measures, and lack of clinician assessment of functional status.Higher levels of pain interference may reflect not only greater physiological impact of the pain problem, but also cognitive factors such as pain catastrophizing. 34While the overall impact of pain on quality of life is multifaceted, we focused on pain interference with daily activities, because of the conceptual similarity between this question and the general measure of disability used in our study, and because additional measures of pain-related limitations were unavailable in the NHIS.Lastly, we were unable to examine detailed clinical history as a predictor of disability presence and type, although we expect a more complete history might have further helped differentiate which people with chronic pain were most likely to report pain interference but not overall limitation in their daily activities.
In conclusion, chronic pain contributes to disability, and the majority of people with both chronic pain and disability endorse activity limitations that are related to pain.However, our study reveals that many people with chronic pain are experiencing interference with daily activities due to pain problems, yet do not respond affirmatively to a question about general disability.Future research should aim to better assess the way people with chronic pain perceive the causative factors in disability and activity limitations, to disentangle the particular contribution of pain problems from disability attributed to other comorbid conditions.While our study has revealed the importance of examining the overlap between dichotomized measures of disability due to pain and disability due to health problems in general, we recommend that future research consider continuous scales for both constructs, to better characterize the contribution of chronic pain to disability burden among respondents who endorse both types of disability in the NHIS and similar surveys.Additionally, our results underscore the importance of addressing pain interference with usual activities, in addition to more general measures of functional status or quality of life, as a treatment goal in rehabilitation therapy for people with chronic pain.

F U N DI NG I N FOR M AT ION
No funding was received for this work.

CON F L IC T OF I N T E R E ST STAT E M E N T
DT discloses salary support for unrelated research and quality improvement projects from the Kate B. Reynolds Charitable Trust and Lilly and Co. Inc.Authors have no other conflicts of interest to report.
Weighted means or proportions (with 95% confidence intervals) of respondent demographic and clinical characteristics, according to the presence of general and pain-related disability.
T A B L E 2 Multivariable multinomial logistic regression model, comparing pain-related but no general disability to both general and pain-related disability.
Abbreviations: CI, confidence interval; Ref., reference group; RRR, relative risk ratio.a Categories not mutually exclusive.