Abstract
Background:
Studies have investigated the association between pain and cognitive impairment among older adults, but the findings are mixed. We assessed the relationship of activity-limiting pain (pain interference) with incident cognitive impairment and the mediating effect of depressive symptoms among Mexican American adults aged ≥80.
Methods:
Data came from the Hispanic Established Population for the Epidemiological Study of the Elderly (2010-2016). Pain interference, or pain that limited daily activities in the last 12 months, was categorized into: none, untreated pain interference, and treated pain interference. Cognitive impairment was defined as scoring <21 on the Mini-Mental State Examination and difficulty with at least one instrumental activity of daily living. We used general estimation equations to assess this relationship between pain and incident cognitive impairment over the six-year period (n=313).
Results:
Participants reporting both untreated and treated pain interference had higher odds of incident cognitive impairment than those reporting no pain or pain interference [untreated adjusted odds ratio (aOR): 2.18; 95% Confidence Interval (CI): 1.09-4.36; treated aOR: 1.99; 95% CI: 1.15-3.44]. Depressive symptoms explained 15.0% of the total effect of untreated pain and 25.3% of treated pain.
Conclusions:
Among very old Mexican American adults, both treated and untreated pain interference was associated with incident cognitive impairment. This association was partially mediated by depressive symptoms, underscoring a need for depression screening in patients with chronic pain. Future work is needed to examine mechanistic/causal pathways between pain and subsequent cognitive impairment and the role of pharmacological and non-pharmacological treatments in these pathways.
Keywords: pain, aging, ethnicity, dementia
Introduction
The prevalence of Alzheimer’s disease and related dementias (ADRD) is projected to double from 2014 to 2060 in adults aged 65 and older.1 Pain, which is common among older adults, may potentially be a modifiable risk factor for dementia. While pain has been associated with worse cognition,2–7 cognitive impairment,7,8 and dementia,2,5,9 some recent studies have not found a relationship between pain and cognitive decline11 or dementia.12 Such mixed findings may be due to varying definitions of pain, cognitive impairment, and dementia. Pain interference, defined as pain that limits a person’s daily activities, may be a more important risk factor for cognitive impairment than pain severity.2,7
Pain treatment is also important to consider, as findings on its relationship to cognition have also been mixed.3,13–15 Many of the epidemiological studies that examined pain and cognitive impairment did not account for pain treatment, due to a lack of available data. The effect of pain treatment may provide insight into the mechanisms of the relationship between pain and cognitive impairment.
The relationship between pain, pain treatment, and cognition, has not been studied among Mexican American older adults, who comprise a majority of the US Hispanic population.16,17 Hispanic adults aged 65 and older have the largest projected increase in the number of ADRD cases, as much as seven-fold from 2014 to 2060.1 Findings on the burden of pain among Mexican American adults are mixed, reflecting differences in the definition of pain (acute vs. chronic, definition of severity, interference with function vs none), in study settings (claims data vs. survey data vs. qualitative studies), and in sample size. Furthermore, few studies take into account the differences in social and cultural determinants of health across various Hispanic groups with different ancestry.18 Previous studies have documented a greater burden of acute pain19 and more severe pain,20 but a lower burden of chronic pain19,21–23 and pain conditions20 among Hispanic and Mexican American adults compared to non-Hispanic White adults. More recent work that disaggregated by Hispanic ancestry documented lower odds of chronic, severe, and debilitating pain among Mexican American adults compared to Non-Hispanic White and Puerto Rican adults.24 Additionally, Hispanic American adults face barriers towards pain treatment including language barriers, financial constraints, suboptimal health insurance, and limited access to specialized pain services.20 Importantly, most of these studies focused on US adults in general rather than older adults.
While consensus does not exist on the mechanism linking pain and cognitive impairment, several hypotheses exist. First, pain competes for cognitive resources and diminishes attentional capability.3,4,33 Additionally, pain affects brain regions involved in cognition, including the prefrontal cortex,34 orbitofrontal cortex,35 and hippocampus,36 regions also involved in depression.37 Inflammation is also hypothesized to play a role in the relationship between pain and cognitive impairment.38 Pain treatment should, theoretically, improve cognition.
The relationship between pain and cognitive impairment may be also be due to shared risk factors, such as depression.25–28 Previous work suggests that depression may be a mediator in the relationship between pain and cognition.9,29,30 Hispanic older adults have a higher prevalence of depressive symptoms compared to non-Hispanic White older adults.31,32 One study using data from the Centers for Medicare and Medicaid Services Health Outcomes Surveys found that 18.0% of Mexican American older adults screened positive for depression using the Patient Health Questionnaire-2, compared to 8.8% of non-Hispanic White participants.31 In fully adjusted models, those identifying as Mexican-origin had 1.19 times the odds of screening positive for depression of their non-Hispanic White counterparts.31 Recent work considering nativity status has documented an increased risk of depressive symptoms among both U.S.-born and foreign-born Hispanic adults (50+), compared to non-Hispanic White adults.32
Our objective is to examine the relationship of pain with incident cognitive impairment and test depressive symptoms as a potential mediator in this relationship, among Mexican American adults aged 80 and older. We hope to clarify the relationship between pain and cognitive impairment by incorporating pain interference and pain treatment into our definition of pain. We hypothesize that pain will be associated with incident cognitive impairment and depressive symptoms will mediate this relationship.
Methods
Sample
Data came from Wave 7 (2010/2011), Wave 8 (2012/2013), and Wave 9 (2016) of the Hispanic Established Population for the Epidemiological Study of the Elderly (H-EPESE), a longitudinal study of Mexican American adults older residing in the Southwestern US.39,40 The H-EPESE started in 1993/1994 with 3,050 subjects, aged 65 years and older, residing in Arizona, California, Colorado, New Mexico, and Texas. The study had a multistage area probability sampling design, including counties that were at least 30% Hispanic.39 Trained bilingual interviewers conducted interviews in the respondent’s home (or their proxy). Follow-up interviews were conducted roughly every two or three years, and at Wave 5 (2004/2005), a probability sample of Mexican American adults aged 75 years and older (n=902) were added to the sample, resulting in a total sample of 2,069 at Wave 5. These individuals were re-interviewed roughly every two or three years, until Wave 9 (2016). Oral informed consent was obtained from all participants and the study protocol was approved by the University of Texas Medical Branch institutional review board (IRB#92-85).
At Wave 7, detailed measures on pain, including pain interference and treatment, were added. Therefore, we used Wave 7 as the baseline of our analysis, with respondents aged 80 and older. Figure 1 describes the analytical sample selection. We excluded those with cognitive impairment at baseline. We also excluded those with missing relevant information at baseline or Wave 8. The final analytic sample included 313 participants. Supplementary Table 1 compares the characteristics of those included in our analytic sample to those excluded. Those excluded were older, female, not married, had fewer years of education, not born in the US, and more often reported health conditions and pain (p<0.05).
Figure 1:
Analytical sample selection. H-EPESE, Hispanic Established Population for the Epidemiological Study of the Elderly.
Measures
Independent variable:
At Waves 7 through 9 participants were asked if, “in the past month, did you notice any pain or discomfort when you stood or walked” (pain on weight-bearing). This question likely captured musculoskeletal pain. Those who reported yes were asked: “in the past month, how much has this pain or discomfort restricted your daily activities— a lot, some or not at all?” (pain interference). Those who reported pain on weight-bearing were also asked if they take any medication for their pain (yes/no). These questions were used as the basis for our pain variable, categorized as no pain interference, untreated pain interference, and treated pain interference. Those who did not report pain or reported non-interfering pain were included in the reference group. Those who reported pain interference but no pain medication use were included in the untreated pain interference group. Lastly, those who reported pain treatment, regardless of pain interference, were included in the treated group, as those who are managing their pain may not report pain interference. We conducted sensitivity analyses for pain on weight-bearing, which may be a less sensitive measure of pain (no pain, untreated pain on weight-bearing, treated pain on weight-bearing). All pain measures included were time-varying, measured at all three waves.
Dependent variable:
Cognitive impairment was defined as <21 on the Mini-Mental State Examination (MMSE)41 and difficulty with at least one instrumental activity of daily living (IADL). A cutoff of 21 has been previously found to be appropriate for samples of older adults with low education and has previously been used with the H-EPESE.42,43 IADL limitations used in this categorization included needing assistance with using the telephone, shopping, preparing meals, taking medicine, and handling money. Cognitive impairment was measured at all three waves.
Potential mediators:
Depressive symptoms were measured at each wave using the Center for Epidemiological Studies Depression (CES-D) Scale.44 The CES-D scale contains 20 questions on frequency of symptoms experienced in the past week, with rarely/none, some/little, occasional/moderate, and most/all of the time as response options [scored from 0 (rarely/none) to 3 (most/all), for a total score ranging 0-60 points]. Participants with a score of 16 or more are considered to have high depressive symptoms. We used the continuous CES-D score in our main models and included the binary categorization of depressive symptoms in sensitivity analyses. Both categorizations of depressive symptoms were time-varying.
Covariates:
Baseline covariates included sex (male/female), years of education, nativity (US born, foreign born). Age, marital status (married, not married), language of interview (English, Spanish), and health conditions (diabetes, stroke, cancer, hypertension, and arthritis) were measured at each wave. Health conditions were assessed by asking participants if they had ever been told by a doctor that they had diabetes/sugar in urine/high blood sugar, stroke/blood clot in the brain/brain hemorrhage, cancer/malignant tumor, high blood pressure, or arthritis/rheumatism.
Statistical analysis
Chi-square tests of independence, Fischer’s exact tests of independence, and analysis of variance (ANOVA) tests were used to assess differences in baseline characteristics by pain status. We used generalized estimating equations (GEE) to assess the relationship of pain with incident cognitive impairment among those with normal cognition at baseline, controlling for sociodemographic and health characteristics. We used a link logit binomial distribution with an autoregressive correlation structure. All models controlled for time in study, age, sex, education, nativity, language of interview, marital status, diabetes, stroke, cancer, hypertension, arthritis, and depressive symptoms. All variables were time-varying, with the potential to change at each wave, except for sex, years of education, and nativity.
We also assessed depressive symptoms as a potential mediator using the Karlson, Holm, and Breen (KHB) method, which decomposes the effect of pain on the odds of cognitive impairment into direct and indirect effects.45 The direct effect is the effect of pain on cognition, not accounting for mediating or confounding variables, while the indirect effect is the portion of the total effect that is explained by mediating variables. We included depressive symptoms (continuous CES-D score) as a potential mediator and included other covariates listed above as potential confounders.
We conducted sensitivity analyses to test the consistency of our results. First, we examined the effect of pain on weight-bearing on incident cognitive impairment. Second, we included high depressive symptoms (CES-D ≥16), as a binary variable, in both the GEE models and the mediation analysis. Lastly, we examined the relationship between pain interference and cognitive impairment, including those with cognitive impairment at baseline. Stata 17.0 (StataCorp, LLC, College Station, TX) was used for all analyses.
Results
Baseline characteristics
Our sample (n=313) was 59.7% female with a mean age of 84.8 years [standard deviation (SD): 3.3; Table 1]. At baseline, in 2010, 39.0% reported treated pain interference, 7.7% reported untreated pain interference, and 53.4% reported no pain interference. Most participants were single, US born, completed their interview in Spanish, and reported hypertension and arthritis. On average, participants reported 5.7 (SD: 4.1) years of education. Those with treated pain more often had arthritis and more depressive symptoms (p<0.05). At Wave 8, (n=313), in 2013, 29.0% of those with treated pain interference, 25.0% of those with untreated pain interference, and 13.9 % of those without pain interference had cognitive impairment (p<0.05; Figure 2). At Wave 9 (n=188), in 2016, 46.9% of those with treated pain interference, 51.6% of those with untreated pain interference, and 28.0% of those with no pain interference had cognitive impairment (p<0.05).
Table 1:
Baseline characteristics of H-EPESE participants included in these analyses, by pain status at baseline (n=313).
| Pain Interference | |||||
|---|---|---|---|---|---|
|
| |||||
| Baseline Characteristic | Total (n=313) |
No (n=167; 53.4%) |
Untreated (n=24; 7.7%) |
Treated (n=122; 39.0%) |
P-Value |
| Age, years (SD) | 84.8 (3.3) | 84.8 (3.4) | 85.3 (4.0) | 84.6 (3.1) | 0.220 |
| Gender | 0.478 | ||||
| Male | 126 (40.3%) | 72 (43.1%) | 10 (41.7%) | 44 (36.1%) | |
| Female | 187 (59.7%) | 95 (56.9%) | 14 (58.3%) | 78 (63.9%) | |
| Marital Status | 0.451 | ||||
| Single | 184 (58.8%) | 103 (61.7%) | 12 (50.0%) | 69 (56.6%) | |
| Married | 129 (41.2%) | 64 (38.3%) | 12 (50.0%) | 53 (43.4%) | |
| Years of Education (SD) | 5.7 (4.1) | 5.9 (4.3) | 6.0 (4.2) | 5.5 (3.7) | 0.278 |
| Nativity | 0.436 | ||||
| US Born | 186 (59.4%) | 104 (62.3%) | 12 (50.0%) | 70 (57.4%) | |
| Mexico Born | 127 (40.6%) | 63 (37.7%) | 12 (50.0%) | 52 (42.6%) | |
| Language of Interview | 0.492 | ||||
| English | 49 (15.7%) | 29 (17.4%) | 2 (8.3%) | 18 (14.8%) | |
| Spanish | 264 (84.3%) | 138 (82.6%) | 22 (91.7%) | 104 (85.2%) | |
| Diabetes | 109 (34.8%) | 52 (31.1%) | 6 (25.0%) | 51 (51.8%) | 0.098 |
| Stroke | 20 (6.4%) | 13 (7.8%) | 2 (8.3%) | 5 (4.1%) | 0.371 |
| Cancer | 34 (10.9%) | 21 (12.6%) | 4 (16.7%) | 9 (7.4%) | 0.215 |
| Hypertension | 237 (75.7%) | 119 (71.3%) | 19 (79.2%) | 99 (81.2%) | 0.141 |
| Arthritis | 203 (64.9%) | 90 (53.9%) | 16 (66.7%) | 97 (79.5%) | <0.001 |
| CES-D score (SD) | 8.2 (7.9) | 6.5 (6.3) | 8.8 (8.5) | 10.5 (9.1) | <0.001 |
| High Depressive Symptoms | 43 (13.7%) | 16 (9.6%) | 3 (12.5%) | 24 (19.7%) | 0.043 |
| MMSE-Score (SD) | 24.7 (4.1) | 25.2 (3.7) | 24.8 (2.9) | 24.0 (4.7) | 0.002 |
| Mean Number of IADL limitations (IQR) | 0.0 (0.0-1.0) | 0.0 (0.0-1.0) | 0.0 (0.0-2.0) | 0.0 (0.0-1.0) | 0.841 |
Note: Some percentages may add up to more than 100% because of rounding. H-EPESE: Hispanic Established Population for the Epidemiological Study of the Elderly; SD: standard deviation; CES-D: Center for Epidemiologic Studies Depression Scale; MMSE: Mini-Mental State Examination; IADL: Instrumental Activities of Daily Living; IQR: Interquartile range.
Figure 2:
Percentage of Mexican American adults aged 80 and older with cognitive impairment by pain interference status in 2012/2013 and 2016.
Generalized Estimating Equations (GEE) models
Table 2 presents the results of our fully adjusted GEE models examining cognitive impairment as a function of pain. Participants contributed 806 observations over the three waves. Those with untreated pain interference had 2.18 times the odds of incident cognitive impairment compared to those who reported no pain or pain interference [95% confidence interval (CI): 1.09-4.36; p=0.028]. Those with treated pain interference had 1.99 times the odds of incident cognitive impairment compared to those who reported no pain or pain interference (95% CI: 1.15-3.44; p=0.014).
Table 2:
Fully adjusted generalized estimating equations for cognitive impairment as a function of pain interference (time-varying), over 6 years among Mexican American adults aged 80+ (n=313).
| Predictor Variables | aOR | 95% CI | p-value |
|---|---|---|---|
| Pain Interference | |||
| None | ref. | ref. | ref. |
| Untreated | 2.18 | 1.09-4.36 | 0.028 |
| Treated | 1.99 | 1.15-3.44 | 0.014 |
| Time (years) | 1.64 | 1.49-1.81 | <0.001 |
| Age (Years) | 1.07 | 1.00-1.15 | 0.066 |
| Sex | |||
| Men | ref. | ref. | ref. |
| Women | 1.56 | 0.86-2.84 | 0.143 |
| Marital Status | |||
| Single | ref. | ref. | ref. |
| Married | 0.96 | 0.55-1.67 | 0.880 |
| Years of Education | 0.78 | 0.71-0.85 | <0.001 |
| Nativity | |||
| Mexico Born | ref. | ref. | ref. |
| US Born | 1.00 | 0.58-1.75 | 0.987 |
| Language of Interview | |||
| English | ref. | ref. | ref. |
| Spanish | 0.99 | 0.43-2.27 | 0.977 |
| Diabetes | 0.63 | 0.35-1.12 | 0.115 |
| Stroke | 2.87 | 1.14-7.21 | 0.025 |
| Cancer | 1.15 | 0.55-2.42 | 0.713 |
| Hypertension | 0.84 | 0.47-1.49 | 0.546 |
| Arthritis | 0.55 | 0.33-0.94 | 0.030 |
| CES-D score | 1.06 | 1.03-1.09 | <0.001 |
Note: aOR: adjusted odds ratio; CI: confidence interval; CES-D: Center for Epidemiologic Studies Depression Scale.
Mediation Analysis
Table 3 displays the summary of our mediation analysis, where we decomposed the effect of pain interference on cognitive impairment into direct and indirect effects via depressive symptoms. The total effect showed that those with untreated pain interference had 2.93 (95% CI: 1.53–5.59) times the odds of incident cognitive impairment and those with treated pain interference had 2.48 (95% CI: 1.50–4.11) times the odds of incident cognitive impairment compared with those with no pain interference. After adjusting for mediating variables, those with untreated pain had 2.50 (95% CI: 1.31–4.78) times the odds of incident cognitive impairment and those with treated pain had 1.98 (95% CI:1.18–3.31) times the odds of incident cognitive impairment. Depressive symptoms partially mediated the relationship between pain interference and incident cognitive impairment. Depressive symptoms explained 15.0% of the total effect for untreated pain interference and 25.3% of the total effect for treated pain interference.
Table 3:
Decomposition of the effect of pain interference (time-varying) on cognitive impairment into direct and indirect effects via depressive symptoms (continuous).
| Mediation Summary |
|||
|---|---|---|---|
| Total Effect | Direct Effect | Indirect Effect | |
| Untreated Pain | |||
| Coefficient (SE) | 1.07 (0.33) | 0.92 (0.33) | 0.16 (0.07) |
| Odds ratio (95% CI) | 2.93 (1.53, 5.59) | 2.50 (1.31, 4.78) | 1.17 (1.02, 1.35) |
| Treated Pain | |||
| Coefficient (SE) | 0.91 (0.26) | 0.68 (0.26) | 0.23 (0.08) |
| Odds ratio (95% CI) | 2.48 (1.50, 4.11) | 1.98 (1.18, 3.31) | 1.25 (1.07, 1.47) |
| Components of the Indirect Effect |
|||
| Coefficient (SE) | % Reduced | p-value | |
|
| |||
| Untreated Pain | |||
| CES-D Score | 0.16 (0.06) | 14.7 | 0.008 |
| Treated Pain | |||
| CES-D Score | 0.23 (0.06) | 24.8 | <0.001 |
Note: The mediating variables explained 15.0% of the total effect for untreated pain and 25.3% of the total effect for treated pain. This percentage was calculated by dividing the indirect effect coefficient by the total effect coefficient. SE: standard error; CI: confidence interval; CES-D: Center for Epidemiologic Studies Depression Scale.
Sensitivity analyses
We conducted sensitivity analyses to test the consistency of our results using varying definitions of pain and depressive symptoms. First, we examined pain on weight-bearing, categorized into no pain on weight-bearing, untreated pain on weight-bearing, and treated pain on weight-bearing (Supplementary Table 2). We found that treated pain on weight-bearing was associated with incident cognitive impairment, but untreated pain on weight-bearing was not. Then, we tested the consistency of our results using high depressive symptoms (yes, no) (Supplementary Table 3). The results of the fully adjusted GEEs were consistent with what was observed in the main models. In the mediation analysis, depressive symptoms were a significant mediator for treated pain interference but not for untreated pain interference. The mediating variables explained 5.6% of the total effect for untreated pain and 15.1% of the total effect for treated pain (Supplementary Table 4). Lastly, we examined the association between pain interference and cognitive impairment, including those with cognitive impairment at baseline (n=424). Findings were similar for those with both untreated and treated pain interference [untreated pain odds ratio (OR): 1.55, 95% CI: 1.10-2.18; treated pain OR: 1.43, 95% CI: 1.09-1.88] (data not shown).
Discussion
In this sample of very old, community-dwelling, Mexican American adults, we found that both untreated pain and treated pain interference were associated with incident cognitive impairment. We also found that depressive symptoms partially mediated this relationship. To our knowledge, this is the first study to investigate the relationship between pain and incident cognitive impairment, examine the potential mediating role of depressive symptoms, consider the degree to which pain interferes with daily activities, and incorporate pain treatment, among very old Mexican American adults.
Our finding that both treated and untreated pain interference was associated with incident cognitive impairment is consistent with studies that did not account for use of pain medication.2,5,7 Ezzati and colleagues found that, among community-dwelling adults aged 70 and older in New York, pain interference, but not pain intensity, was associated with incident dementia.2 Bell and colleagues found that, among adults aged 65 and older in the US, pain presence and pain interference, but not pain intensity, were associated with odds of being classified as cognitively impaired.7 In our sensitivity analysis, where we examined pain on weight-bearing, only treated pain was associated with cognitive impairment, suggesting either more severe pain needing treatment or the possibility of pain treatment associating with cognitive change. Pain on weight-bearing is a less sensitive pain measure than pain interference, and those with untreated pain on weight-bearing might have milder pain than those who require treatment.
Previous work that did account for analgesic use found that pain was associated not with major cognitive impairment but with impairment in memory domains.8 Other work documented a significant relationship between non-cancer chronic pain conditions and ADRD, even after controlling for analgesic use.9 Although we observed similar findings among those with treated and untreated pain, our untreated group was small (n=25), indicating the need for future studies to understand the effect of type and duration of pain treatments on subsequent change in cognition in patients with chronic pain. It is unclear if pain treatment contributes to cognitive impairment or alleviates the cognitive impairment associated with pain.3 Findings on opioid use and the risk of dementia among older adults are mixed. One study documented that older adults with the heaviest opioid use [91+ total standardized doses (TSD; 1 TSD equals 30 mg of morphine)] had a 29% higher risk of dementia than those with 0-10 TSD.46 However, this study lacked information on pain, so this greater dementia risk may be due to residual chronic pain, rather than the pain treatment itself.46 Other studies found no relationship between opioid use and cognitive decline47 or ADRD.15
Few studies have assessed the potential mediating role of depression or depressive symptomatology. One study found that sleep disturbances and depressive symptoms fully mediated the relationship between pain and cognitive dysfunction among adults aged 18 and older living with systemic lupus erythematosus.30 Another study found that depression fully mediated the relationship between pain and cognition among community-dwelling patients with rheumatoid arthritis, aged 34 to 84.29 The full mediation noted by these two studies may reflect their examination of patient populations with an autoimmune disorder. Recent work analyzing data from community-dwelling Medicare beneficiaries aged 65 and older found that non-cancer chronic pain increased the risk of Alzheimer’s disease, and that mood disorders (depression/anxiety) partially mediated this relationship.9 This study and our current findings underscore a need for routine screening for depressive symptoms in patients with chronic pain, especially in the context of interference with daily activities. Whether treating depression will reduce the likelihood of a pain-cognitive impairment association is unclear. Screening and treatment for depressive symptoms are especially important in very old Mexican American adults, since both structural and cultural barriers may limit their access to care.48
This work may have clinical implications for older adults living with pain; however, future work is needed to understand the role of pain treatment in the relationship between pain and cognitive impairment. This is a critical gap in the current literature. If pain treatment helps alleviate the effects of pain on cognitive function, then treatment represents an important potentially modifiable risk factor for cognitive impairment. This is particularly relevant to Hispanic older adults who may delay seeking pain treatment due to language barriers and suboptimal access to health care.20 Additionally, Hispanic adults may prefer to not take pain medication because of stoicism, fear of addiction, or religious beliefs.49 A qualitative study conducted among foreign-born Hispanic patients at a Federally Qualified Health Center noted that participants preferred alternatives to pain medication, such as home remedies, and that seeking pain management from a doctor was their last resort.50 This study also found a low awareness of non-pharmacological pain treatments (physical therapy, surgery, psychosocial treatment) but an interest in learning more about these options.50 Non-pharmacological pain treatments may represent an opportunity to improve pain management in older Hispanic adults. Future work is warranted on the influence of pain management on cognitive impairment.
Limitations
Our study has certain limitations. First, selection bias may have occurred through the exclusion of those with cognitive impairment at baseline and the inclusion of individuals aged 80 and older. We also required that participants have at least two observations, which may have introduced selection bias as participants who died after baseline were not included. This may have resulted in a more robust, healthier, sample, and may have underestimated the true relationship between pain and cognitive impairment. However, in sensitivity analysis including those with cognitive impairment at baseline, we see similar findings. Second, our pain and pain medication measures are limited, do not have detailed information on participant’s pain experience, only collect information on pain in the last month, and do not have detailed treatment information. Third, our findings may be subject to recall bias, due to our subjects’ advanced age. Fourth, our findings may be not generalizable to the US Hispanic population as a whole, as the US Hispanic population is heterogeneous and pain burden may differ by country of origin.24 For instance, a recent paper found that those of Puerto Rican origin report more chronic and severe pain, while those of Mexican origin report less chronic and severe pain.24 Fifth, our small sample size, particularly among those with untreated pain, may have limited our ability to see nuanced differences in cognition by pain status. Lastly, we are not able to establish the exact temporal relationship between pain, depressive symptoms, and cognition due to the observational nature of our data. We cannot determine if pain preceded the onset of depressive symptoms. Study strengths include the examination of pain medication and the use of a well-categorized, unique population.
Conclusions
In our sample of very old, community-dwelling, Mexican American adults, both untreated and treated pain interference were associated with incident cognitive impairment. Depressive symptoms also partially mediated this relationship. Pain and depressive symptoms may be modifiable risk factors for cognitive function. Options for pain and depression treatment exist and may improve cognitive outcomes of older adults with activity-limiting pain. Future work should address the biopsychosocial determinants of undertreatment of pain in very old Mexican American populations and examine causal mechanisms between pain, depressive symptoms, and cognitive impairment, to provide data to guide clinical practice and inform policy on pain treatment. Future work should also examine the role of non-pharmacological treatments in the relationship between pain and cognitive impairment.
Supplementary Material
Supplementary Table 1: Baseline characteristics of H-EPESE participants by inclusion in the analytical sample.
Supplementary Table 2: Fully adjusted generalized estimating equations for cognitive impairment as a function of pain on weight-bearing, over 6 years among Mexican Americans aged 80+ (n=323).
Supplementary Table 3: Fully adjusted generalized estimating equations for cognitive impairment as a function of pain interference, controlling for elevated depressive symptoms (binary), over 6 years among Mexican Americans aged 80+ (n=323).
Supplementary Table 4: Decomposition of the effect of pain on weight-bearing on cognitive impairment into direct and indirect effects via pain medication and elevated depressive symptoms (binary) score.
Key Points:
Both treated and untreated pain interference were associated with incident cognitive impairment over six-years among Mexican American adults aged 80 and older.
Depressive symptoms partially mediated the relationship between pain and incident cognitive impairment.
Why does this matter?
Pain and depressive symptoms may be potentially modifiable risk factors for incident cognitive impairment.
Disclosures:
This work was supported by the National Institute on Aging (grant numbers K01AG075254, R01AG10939, P30AG024832, P30AG059301, and RF1AG068988), the National Institute on Minority Health and Health Disparities (grant number R01MD010355), and the National Institute on Drug Abuse (R01DA039192). Dr. S. A. Milani was supported by a research career development award (K12HD052023: Building Interdisciplinary Research Careers in Women’s Health Program-BIRCWH; Berenson, PI) from the National Institutes of Health/Office of the Director (OD)/National Institute of Allergy and Infectious Diseases (NIAID), and the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD).
Footnotes
Conflicts of interest: None reported.
References
- 1.Matthews KA, Xu W, Gaglioti AH, et al. Racial and ethnic estimates of Alzheimer’s disease and related dementias in the United States (2015-2060) in adults aged ≥65 years. Alzheimer’s & dementia : the journal of the Alzheimer’s Association. 2019;15(1):17–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Ezzati A, Wang C, Katz MJ, et al. The Temporal Relationship between Pain Intensity and Pain Interference and Incident Dementia. Curr Alzheimer Res. 2019;16(2):109–115. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Moriarty O, McGuire BE, Finn DP. The effect of pain on cognitive function: a review of clinical and preclinical research. Prog Neurobiol. 2011;93(3):385–404. [DOI] [PubMed] [Google Scholar]
- 4.van der Leeuw G, Eggermont LH, Shi L, et al. Pain and Cognitive Function Among Older Adults Living in the Community. The journals of gerontology Series A, Biological sciences and medical sciences. 2016;71(3):398–405. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Whitlock EL, Diaz-Ramirez LG, Glymour MM, Boscardin WJ, Covinsky KE, Smith AK. Association Between Persistent Pain and Memory Decline and Dementia in a Longitudinal Cohort of Elders. JAMA internal medicine. 2017;177(8):1146–1153. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Milani SA, Bell TR, Crowe M, Pope CN, Downer B. Increasing Pain Interference is Associated with Cognitive Decline over Four Years among Older Puerto Rican Adults. The journals of gerontology Series A, Biological sciences and medical sciences. 2022. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Bell T, Franz CE, Kremen WS. Persistence of pain and cognitive impairment in older adults. J Am Geriatr Soc. 2022;n/a(n/a). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.van der Leeuw G, Ayers E, Leveille SG, Blankenstein AH, van der Horst HE, Verghese J. The Effect of Pain on Major Cognitive Impairment in Older Adults. The Journal of Pain. 2018;19(12):1435–1444. [DOI] [PubMed] [Google Scholar]
- 9.Khalid S, Sambamoorthi U, Innes KE. Non-Cancer Chronic Pain Conditions and Risk for Incident Alzheimer’s Disease and Related Dementias in Community-Dwelling Older Adults: A Population-Based Retrospective Cohort Study of United States Medicare Beneficiaries, 2001-2013. Int J Environ Res Public Health. 2020;17(15). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Kumaradev S, Fayosse A, Dugravot A, et al. Timeline of pain before dementia diagnosis: a 27-year follow-up study. Pain. 2021;162(5):1578–1585. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.de Aguiar G, Saraiva MD, Khazaal EJB, de Andrade DC, Jacob-Filho W, Suemoto CK. Persistent pain and cognitive decline in older adults: a systematic review and meta-analysis from longitudinal studies. Pain. 2020;161(10):2236–2247. [DOI] [PubMed] [Google Scholar]
- 12.Rouch I, Edjolo A, Laurent B, Dartigues JF, Amieva H. Chronic pain and long-term dementia risk in older adults: Results from a 24-year longitudinal study. Int J Geriatr Psychiatry. 2022;37(5). [DOI] [PubMed] [Google Scholar]
- 13.Pritchard KT, Baillargeon J, Raji MA, Chou L-N, Downer B, Kuo Y-F. Association of Occupational and Physical Therapy With Duration of Prescription Opioid Use After Hip or Knee Arthroplasty: A Retrospective Cohort Study of Medicare Enrollees. Archives of Physical Medicine and Rehabilitation. 2021;102(7):1257–1266. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Dublin S, Walker RL, Gray SL, et al. Prescription Opioids and Risk of Dementia or Cognitive Decline: A Prospective Cohort Study. J Am Geriatr Soc. 2015;63(8):1519–1526. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Taipale H, Hamina A, Lampela P, et al. Is Alzheimer’s Disease Associated with Previous Opioid Use? Pain Medicine. 2018;19(11):2115–2121. [DOI] [PubMed] [Google Scholar]
- 16.Frey WH. The US will become ‘minority white’ in 2045, Census projects: Youthful minorities are the engine of future growth. In:2018. [Google Scholar]
- 17.Noe-Bustamante L Key facts about U.S. Hispanics and their diverse heritage. https://www.pewresearch.org/fact-tank/2019/09/16/key-facts-about-u-s-hispanics/. Published 2019. Accessed November 11, 2020.
- 18.Garcia MA, Warner DF, García C, Downer B, Raji M. Age Patterns in Self-Reported Cognitive Impairment Among Older Latino Subgroups and Non-Latino Whites in the United States, 1997-2018: Implications for Public Health Policy. Innov Aging. 2021;5(4):igab039. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Riskowski JL. Associations of Socioeconomic Position and Pain Prevalence in the United States: Findings from the National Health and Nutrition Examination Survey. Pain Medicine. 2014;15(9):1508–1521. [DOI] [PubMed] [Google Scholar]
- 20.Hollingshead NA, Ashburn-Nardo L, Stewart JC, Hirsh AT. The Pain Experience of Hispanic Americans: A Critical Literature Review and Conceptual Model. The Journal of Pain. 2016;17(5):513–528. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Hollingshead NA, Vrany EA, Stewart JC, Hirsh AT. Differences in Mexican Americans’ Prevalence of Chronic Pain and Co-Occurring Analgesic Medication and Substance Use Relative to Non-Hispanic White and Black Americans: Results from NHANES 1999–2004. Pain Medicine. 2015;17(6):1001–1009. [DOI] [PubMed] [Google Scholar]
- 22.Hardt J, Jacobsen C, Goldberg J, Nickel R, Buchwald D. Prevalence of Chronic Pain in a Representative Sample in the United States. Pain Medicine. 2008;9(7):803–812. [DOI] [PubMed] [Google Scholar]
- 23.Zajacova A, Grol-Prokopczyk H, Fillingim R. Beyond Black vs White: racial/ethnic disparities in chronic pain including Hispanic, Asian, Native American, and multiracial US adults. Pain. 2022;163(9). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Nahin RL. Pain Prevalence, Chronicity and Impact Within Subpopulations Based on Both Hispanic Ancestry and Race: United States, 2010-2017. The Journal of Pain. 2021. [DOI] [PubMed] [Google Scholar]
- 25.Zis P, Daskalaki A, Bountouni I, Sykioti P, Varrassi G, Paladini A. Depression and chronic pain in the elderly: links and management challenges. Clinical interventions in aging. 2017;12:709–720. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Gerrits MMJG, van Marwijk HWJ, van Oppen P, van der Horst H, Penninx BWJH. Longitudinal association between pain, and depression and anxiety over four years. Journal of Psychosomatic Research. 2015;78(1):64–70. [DOI] [PubMed] [Google Scholar]
- 27.Raji MA, Reyes-Ortiz CA, Kuo Y-F, Markides KS, Ottenbacher KJ. Depressive Symptoms and Cognitive Change in Older Mexican Americans. Journal of geriatric psychiatry and neurology. 2007;20(3):145–152. [DOI] [PubMed] [Google Scholar]
- 28.Thomas AJ, O’Brien JT. Depression and cognition in older adults. Current Opinion in Psychiatry. 2008;21(1). [DOI] [PubMed] [Google Scholar]
- 29.Brown SC, Glass JM, Park DC. The relationship of pain and depression to cognitive function in rheumatoid arthritis patients. Pain. 2002;96(3). [DOI] [PubMed] [Google Scholar]
- 30.Lillis TA, Tirone V, Gandhi N, et al. Sleep Disturbance and Depression Symptoms Mediate Relationship Between Pain and Cognitive Dysfunction in Lupus. Arthritis Care Res (Hoboken). 2019;71(3):406–412. [DOI] [PubMed] [Google Scholar]
- 31.Hooker K, Phibbs S, Irvin VL, et al. Depression Among Older Adults in the United States by Disaggregated Race and Ethnicity. Gerontologist. 2019;59(5):886–891. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Boen CE, Hummer RA. Longer-but Harder-Lives?: The Hispanic Health Paradox and the Social Determinants of Racial, Ethnic, and Immigrant-Native Health Disparities from Midlife through Late Life. J Health Soc Behav. 2019;60(4):434–452. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Eccleston C, Crombez G. Pain demands attention: a cognitive-affective model of the interruptive function of pain. Psychol Bull. 1999;125(3):356–366. [DOI] [PubMed] [Google Scholar]
- 34.Miller EK. The prefrontal cortex and cognitive control. Nat Rev Neurosci. 2000;1(1):59–65. [DOI] [PubMed] [Google Scholar]
- 35.Mesulam MM. From sensation to cognition. Brain. 1998;121(6):1013–1052. [DOI] [PubMed] [Google Scholar]
- 36.Zimmerman ME, Pan JW, Hetherington HP, Lipton ML, Baigi K, Lipton RB. Hippocampal correlates of pain in healthy elderly adults: a pilot study. Neurology. 2009;73(19):1567–1570. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Sheng J, Liu S, Wang Y, Cui R, Zhang X. The Link between Depression and Chronic Pain: Neural Mechanisms in the Brain. Neural Plast. 2017;2017:9724371–9724371. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Cao S, Fisher DW, Yu T, Dong H. The link between chronic pain and Alzheimer’s disease. Journal of Neuroinflammation. 2019;16(1):204. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Cantu P, Markides K. Hispanic EPESE (Established Population for the Epidemiological Study of the Elderly). In: Gu D, Dupre ME, eds. Encyclopedia of Gerontology and Population Aging. Cham: Springer International Publishing; 2019:1–8. [Google Scholar]
- 40.Markides K, Rudkin L, Angel R, Espino D. Health status of Hispanic elderly in the United States. En Martin L, Soldo B (Eds). Racial and Ethnic Differences in the Health of Older Americans. In: Washington: National Academy Press; 1997. [Google Scholar]
- 41.Folstein MF, Folstein SE, McHugh PR. “Mini-mental state”. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res. 1975;12(3):189–198. [DOI] [PubMed] [Google Scholar]
- 42.Raji MA, Al Snih S, Ray LA, Patel KV, Markides KS. Cognitive status and incident disability in older Mexican Americans: findings from the Hispanic established population for the epidemiological study of the elderly. Ethn Dis. 2004;14(1):26–31. [PubMed] [Google Scholar]
- 43.Downer B, Al Snih S, Howrey BT, Raji MA, Markides KS, Ottenbacher KJ. Combined effects of cognitive impairment and pre-frailty on future frailty and death in older Mexican Americans. Aging & Mental Health. 2019;23(10):1405–1412. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Radloff LS. The CES-D scale: A self-report depression scale for research in the general population. Applied psychological measurement. 1977;1(3):385–401. [Google Scholar]
- 45.Kohler U, Karlson KB, Holm A. Comparing coefficients of nested nonlinear probability models. The Stata Journal. 2011;11(3):420–438. [Google Scholar]
- 46.Dublin S, Walker RL, Gray SL, et al. Prescription Opioids and Risk of Dementia or Cognitive Decline: A Prospective Cohort Study. J Am Geriatr Soc. 2015;63(8):1519–1526. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Pritchard KT, Downer B, Raji MA, Baillargeon J, Kuo YF. Incident Functional Limitations Among Community-Dwelling Adults Using Opioids: A Retrospective Cohort Study Using a Propensity Analysis with the Health and Retirement Study. Drugs Aging. 2022;39(7):559–571. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Jimenez DE, Martinez Garza D, Cárdenas V, Marquine M. Older Latino Mental Health: A Complicated Picture. Innov Aging. 2020;4(5):igaa033. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Campbell LC, Andrews N, Scipio C, Flores B, Feliu MH, Keefe FJ. Pain Coping in Latino Populations. The Journal of Pain. 2009;10(10):1012–1019. [DOI] [PubMed] [Google Scholar]
- 50.Torres CA, Thorn BE, Kapoor S, DeMonte C. An Examination of Cultural Values and Pain Management in Foreign-Born Spanish-Speaking Hispanics Seeking Care at a Federally Qualified Health Center. Pain Medicine. 2017;18(11):2058–2069. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplementary Table 1: Baseline characteristics of H-EPESE participants by inclusion in the analytical sample.
Supplementary Table 2: Fully adjusted generalized estimating equations for cognitive impairment as a function of pain on weight-bearing, over 6 years among Mexican Americans aged 80+ (n=323).
Supplementary Table 3: Fully adjusted generalized estimating equations for cognitive impairment as a function of pain interference, controlling for elevated depressive symptoms (binary), over 6 years among Mexican Americans aged 80+ (n=323).
Supplementary Table 4: Decomposition of the effect of pain on weight-bearing on cognitive impairment into direct and indirect effects via pain medication and elevated depressive symptoms (binary) score.