ADVERTISEMENT

Issue Navigator

Volume 14 No. 10
Earn CME
Accepted Papers





Scientific Investigations

Chronic Pain, Sleep, and Cognition in Older Adults With Insomnia: A Daily Multilevel Analysis

Ashley F. Curtis, PhD1; Jacob M. Williams, PhD2; Karin J.M. McCoy, PhD3; Christina S. McCrae, PhD1
1Department of Psychiatry, University of Missouri-Columbia, Columbia, Missouri; 2TIRR Memorial Hermann, Houston, Texas; 3Neuropsychology Service, South Texas Veterans Health Care System, San Antonio, Texas

ABSTRACT

Study Objectives:

The goal of this study was to examine daily associations between sleep and cognition in older adults suffering from insomnia, with or without a history of chronic pain.

Methods:

Sixty older adults with insomnia and a history of chronic pain (HxCP; n = 33, mean age = 69.5 years, standard deviation = 7.8) or no history of chronic pain (NCP; n = 27, mean age = 69.7 years, standard deviation = 7.9) completed 14 days of diaries and actigraphy, measuring sleep onset latency (SOL), wake after sleep onset (WASO), sleep efficiency (SE), and sleep quality. Participants completed daily cognitive measures of processing speed (ie, symbol digit modalities test, SDMT), reasoning (ie, letter series), and verbal memory (ie, word list delayed recall). For HxCP and NCP, associations between sleep parameters, daily pain, depressive symptoms (ie, Beck Depression Inventory, Second Edition scores), and daily cognition, controlling for age, and global cognition were examined through multilevel modeling.

Results:

For HxCP, greater self-reported WASO was associated with worse next-day SDMT performance, whereas greater actigraphic WASO was associated with better next-day SDMT performance. Greater depression was associated with worse daily letter series performance. Greater self-reported WASO and SE were associated with better next-day delayed recall. For NCP, greater self-reported WASO and depression were associated with better daily SDMT performance, whereas worse daily pain was associated with worse SDMT and delayed recall performance.

Conclusions:

In older adults with HxCP, improving sleep may benefit lower level cognition, whereas reducing depression may affect higher level cognition. Discrepancies in sleep parameters promote assessment of objective and subjective sleep outcomes when investigating effects of insomnia on cognition.

Clinical Trial Registration:

Title: Intraindividual Variability in Sleep and Cognitive Performance in Older Adults (REST), Registry: ClinicalTrials.gov, Identifier: NCT02967185, URL: https://clinicaltrials.gov/ct2/show/NCT02967185

Citation:

Curtis AF, Williams JM, McCoy KJ, McCrae CS. Chronic pain, sleep, and cognition in older adults with insomnia: a daily multilevel analysis. J Clin Sleep Med. 2018;14(10):1765–1772.


BRIEF SUMMARY

Current Knowledge/Study Rationale: Both insomnia and chronic pain have been associated with worse cognitive performance. However, these conditions have not been examined together in older adults, and it is currently unclear how pain and sleep affect day to day cognitive performance.

Study Impact: Results suggest that the patterns of association between sleep, mood, daily pain, and next-day cognitive performance differ for older adults with and without a history of chronic pain. History of chronic pain should be considered in the management of cognitive disturbances in older adults with insomnia.

INTRODUCTION

Insomnia and chronic pain conditions (eg, fibromyalgia, rheumatoid arthritis, chronic lower back pain) affect up to 50%1 and 31%2 of older adults, respectively. Among older adults with insomnia, up to 50% also experience chronic pain.3 These conditions place an enormous burden on quality of life,4 but their association with cognitive performance in older adults, who are already vulnerable to age-related cognitive decline,5 is less clear. Thus, understanding the extent of these relationships in older adults, as well as the underlying factors or mechanisms associated with cognitive disturbances is warranted.

Investigations of poor sleep and cognition have yielded inconsistent results. For instance, one study on older adults found that insomnia is associated with poor verbal memory performance,6 whereas other studies found no relationship between insomnia and cognitive function.7,8 A study examining sleep quality and cognitive performance in older adults with insomnia found that poor sleep quality was associated with worse performance on tests measuring on executive functions (eg, working memory, attentional set shifting, abstract reasoning), but was not associated with episodic memory or processing speed.9 Interestingly, we also previously reported results showing that longer total wake time was associated with better performance on a sustained attention and processing speed task,10 suggesting that increases in nighttime wakefulness may translate to a state of hyperarousal that provides a benefit to lower level cognitive processing.11 However, these findings represented aggregate patterns of average performance across 2 weeks rather than daily patterns of association and did not consider the effects of chronic pain. Examining the daily associations between nighttime sleep and next-day cognition in the context of a history of chronic pain or no history of chronic pain would provide more meaningful data that could facilitate interventions aimed at improving sleep and mitigating daily cognitive fluctuations.

Results regarding associations between chronic pain and cognition are more consistent, as chronic pain in older adults has been associated with worse working memory performance,12 and worse performance across a wide range of cognitive tasks.13 However, the relationship between insomnia, chronic pain, and cognition in older adults has not been established. That is, it is unclear whether older adults with insomnia and a history of chronic pain differ in the patterns of association between sleep and cognitive performance relative to older adults with insomnia and no history of chronic pain. Furthermore, given that mood disturbance such as increased depression is highly prevalent among individuals with chronic pain conditions,14 and depression has been associated with worse performance across a wide range of cognitive tasks,15 it is important to examine whether any changes in cognitive performance among older adults with insomnia and chronic pain may be associated with fluctuations in mood. Understanding the factors (eg, sleep, pain ratings, mood) that contribute to day-to-day cognitive changes in these individuals may point to potential targets for intervention.

In the current study our aim was to examine the daily associations between nighttime sleep, and daily cognitive performance in older adults with and without a history of chronic pain. We predicted that poor sleep (ie, longer sleep onset, more time spent awake after sleep onset, lower sleep efficiency) would be associated with worse cognitive performance, and this association would be strongest in individuals with a history of chronic pain. We also explored whether daily pain ratings and average mood ratings had an effect on daily cognitive performance. We predicted that although both of these would have a negative association with cognition, this relationship would be strongest for older adults with insomnia and a history of chronic pain.

METHODS

Participants

Participants were recruited as part of a parent clinical trial (NCT02967185 registered at clinicaltrials.gov) that investigated the effects of brief cognitive behavioral therapy on reducing insomnia symptoms in older adults. All participants provided informed consent. The University of Florida Institutional Review Board approved all procedures. Participants were included if they were 65+ years of age, and reported insomnia symptoms (ie, sleep onset or awake time during night > 30 minutes at least 3 nights per week for more than 6 months), reported daytime dysfunction due to insomnia), and reported no prescribed or over-the-counter sleep medication for at least 1 month, or were stabilized on medication for at least 6 months. Exclusion criteria were: sleep disorders other than insomnia (eg, specifically sleep apnea, periodic limb movements disorder), significant medical (eg, cancer) or neurological (eg, dementia) conditions; severe untreated psychiatric comorbidity (eg, schizophrenia, substance abuse); severe depressive symptomatology (ie, a score of 24 or higher on the Beck Depression Inventory, Second Edition (BDI-II)16 or a score of 13 or higher on the Geriatric Depression Scale17); cognitive impairment based on Mini-Mental State Examination (MMSE) score below 23 (for individuals with 9th grade education or higher) or below 18 (for individuals with less than 9th grade education),18,19 use of any psychotropic or other medications (eg, beta blockers) known to alter sleep. The current study examines data from participants who completed the 14 days of baseline prior to treatment.

Sleep Parameters

Sleep Diaries

Participants completed sleep diaries20 upon awakening each morning for the 14 days of baseline. Sleep diaries provided the following daily values: sleep onset latency (SOL) in minutes; total time spent awake (in minutes) after sleep onset (WASO); sleep efficiency percentage: computed as the ratio of total time spent asleep to total time spent in bed × 100%; and sleep quality rating: rated on a scale from 1 (very poor) to 5 (excellent). Additional details on sleep diaries are provided in the previously published clinical trial outcomes manuscript.21

Actigraphy

Subjects wore an Actiwatch-L (ACTL) with an ambient light sensor (Mini Mitter Co., Inc.) on their nondominant wrists 24 hours per day for the entire 14 days of baseline. The actiwatch monitors ambient light exposure and gross motor activity at a rate of 32 samples per second, and records peak values for each second. These peak values are then summed into 30-second “activity” counts, and analyzed using Actiware-Sleep v. 3.3, which uses validated algorithms to classify each epoch as sleep or wake. The scoring period was set to diary reported initial bedtime and final waketime. The following outcomes were computed from actigraphy assessments; SOL, WASO, and sleep efficiency.

Pain Variables

History of Chronic Pain

As part of the participant screening, participants were required to state whether they did or did not have a history of chronic pain. Participants with a history of chronic pain were asked to specify which type (ie, none, arthritis, chronic back pain, hip pain, back and hips, osteoporosis, fibromyalgia, or other).

Pain Ratings

As part of their daily diaries, participants were instructed to provide subjective ratings of their pain level in the morning. As recommended in a consensus statement regarding chronic pain criteria,22 participants rated their pain on a 0 (No Pain) to 10 (Worst Possible Pain) scale.

Cognitive Measures

As part of their daily diaries, participants were asked to complete three cognitive tests, described in detail in the next paragraphs. Participants were instructed to sign, date, and provide the time of completion for each test. To reduce effects of practice, alternate forms were constructed (of equal difficulty) for each of the 14 days. Participants were provided with a stopwatch to use on the timed sections of the tests.

Symbol Digit Modalities Test

The Symbol Digit Modalities Test (SDMT)23 measures certain aspects of attention (ie, selective and sustained) and processing speed.24 The test consists of a legend of nine digit and symbol pairs that is provided at the top of a page of paper, and a series of symbols with blank spaces underneath on the rest of the page. For each symbol, participants are instructed to record the corresponding number (provided in the legend). They are to complete as many number entries as possible within 90 seconds (participants time themselves using a stopwatch). The total score is the number of correct responses (ie, correct numbers that correspond with the digit) provided in the 90-second time limit. Therefore, higher scores correspond to better performance.

Letter Series

The letter series test25 measures inductive reasoning ability. In this test, participants are presented with 30 series of letters and are instructed to choose the letter (from five possible choices) that would continue the established pattern. Participants are instructed to complete as many of the 30 items as possible in 4 minutes (timing themselves using a stopwatch). The total score is the number of correct items provided during this time period. Therefore, higher scores correspond to better performance.

Word List Delayed Recall

The word list delayed recall test was adopted from Rey Auditory Verbal Learning Test26 and measured episodic verbal memory. Participants completed 3 trials in which they study the same list of 15 words for 90 seconds (self-timed), and write down as many of the words as they can remember in 60 seconds (self-timed) without looking at the original list. After the third trial, participants complete a series of other cognitive tasks (ie, SDMT, letter series test) and questionnaires (eg, pain ratings). Then on the delayed recall trial, participants are asked to recall as many words as they can remember in 60 seconds (self-timed) from the original studied list. The delayed recall score is the total number of correct words remembered. Therefore, higher scores correspond to better memory performance.

Mood Measure

Beck Depression Inventory, Second Edition

The BDI-II16 is a 21-item self-report questionnaire that measures the severity of depressive symptoms on a three-point scale [0 (absence of symptoms) to 3 (most severe)]. The scale requires participants to rate how they have been feeling over the past week, but in order to match the recording period at baseline, participants were asked to provide ratings corresponding to the past 2 weeks. Scores range from 0 to 63 and can be classified as follows: under 10 = none or minimal depression, 10 to 18 = mild to moderate depression, 19 to 29 = moderate to severe depression, and over 29 = severe depression.

Statistical Analysis

Differences between the two participant groups (history of chronic pain versus no history of chronic pain) were analyzed using independent sample t tests for continuous variables (age, insomnia duration, pain ratings, MMSE, BDI-II, sleep diary, actigraphy, and cognitive measures), and chi square analysis for categorical variables (sex, education, depression severity, pain rating category, use of prescription sleep medication). Multilevel modeling (MLM) analyses were conducted in SPSS (v.24). For each daily cognitive measure (SDMT, letter series, word list delayed recall), we investigated the within-person fixed effects of daily sleep diary (ie, SOL, WASO, sleep efficiency, sleep quality) and actigraphy (SOL, WASO, sleep efficiency) measured sleep parameters, as well as between-person depressive symptoms (ie, BDI-II score) and within-person self-reported daily pain. Analyses controlled for between-person variability in age and global cognition (ie, MMSE score). Within-person variables were person mean-centered to capture within-person variability, and between-person variables were grand mean-centered to capture between-person variability.27 Across models, the most parsimonious random structure based on goodness of fit was used.28 Factors and covariates were added to the random intercept of each model, and models were evaluated through maximum likelihood estimation.

RESULTS

Participant Characteristics

Participant demographics and average sleep outcomes are provided in Table 1. Participants with a history of chronic pain reported the following conditions: arthritis (n = 25, 76%), back pain (n = 3, 9%), osteoporosis (n = 2, 6%), arthritis and back pain (n = 1, 3%), back and hip pain (n = 1, 3%), and fibromyalgia (n = 1, 3%). Participants with a history of chronic pain had higher average pain ratings relative to participants without a history of chronic pain. The two groups did not differ on any of the other demographic or average sleep, or cognitive outcomes.

Demographics, average sleep parameters and cognitive performance of older adults with and without a history of chronic pain (n = 60).

jcsm.14.10.1765.t01.jpg

table icon
Table 1

Demographics, average sleep parameters and cognitive performance of older adults with and without a history of chronic pain (n = 60).

(more ...)

Multilevel Modelling Results

The fixed effects of diary and actigraphy-measured sleep parameters on daily cognitive performance outcomes are presented in Table 2.

Fixed effects of daily self-report and actigraphy measured sleep parameters on daily cognitive performance in older adults with insomnia with and without a history of chronic pain (n = 60).

jcsm.14.10.1765.t02.jpg

table icon
Table 2

Fixed effects of daily self-report and actigraphy measured sleep parameters on daily cognitive performance in older adults with insomnia with and without a history of chronic pain (n = 60).

(more ...)

Symbol Digit Modalities Test

For participants with a history of chronic pain, greater self-reported WASO was associated with worse next-day cognitive performance, whereas greater actigraphic estimates of WASO were associated with better next-day cognitive performance. None of the other nighttime sleep parameters or other variables (age, daily pain ratings, mood ratings, and global cognition) were associated with daily SDMT performance.

For participants without a history of chronic pain, greater self-reported WASO was associated with better next-day SDMT performance. Better overall global cognition and more depressive symptoms were also associated with better daily performance on the SDMT. Finally, higher daily pain ratings and older age were associated with worse cognitive performance.

Letter Series

For participants with a history of chronic pain, older age and better overall global cognition were associated with better performance on the letter series test. More depressive symptoms were associated with worse performance on the letter series test. None of the sleep variables were associated with next-day letter series performance.

For participants without a history of chronic pain, higher MMSE scores were associated with better letter series performance. Sleep, pain, mood, or age variables showed no association with daily performance.

Word List Delayed Recall

For participants with a history of chronic pain, greater nighttime self-reported WASO was associated with better next-day performance on word list delayed recall. Better self-reported sleep efficiency was associated with better delayed recall. Older age and higher global cognition scores were associated with better delayed recall performance. Effects of all other variables were nonsignificant.

For participants without a history of chronic pain, older age was associated with worse performance on word list delayed recall, and higher global cognition was associated with better performance. Higher morning pain ratings were associated with worse delayed recall. None of the other examined variables showed an association with recall performance.

DISCUSSION

The current study examined whether older adults with insomnia and either a history or no history of chronic pain differed in their patterns of association between daily nighttime sleep parameters, morning pain ratings, and average mood with dayto-day cognitive performance. Patterns of association differed for participants with and without a history of chronic pain, and across cognitive outcomes.

Our prediction that worse sleep would impair next-day cognitive performance, particularly in those older adults with a history of chronic pain was partially supported. In participants with a history of chronic pain, the expected negative association between WASO and SDMT performance was observed. That is, worse self-reported sleep disturbance (ie, increased WASO) was associated with worse next-day cognitive performance. Although these results suggest that poor sleep may negatively affect next-day lower level cognitive performance such as that observed on a selective attention and processing speed task, we note that we observed the opposite association (ie, a positive association between WASO and SDMT performance) for actigraphically measured sleep. Although discrepancies between subjective and objectively measured sleep have been reported in older adults with insomnia,29 and this may reflect differences in the underlying construct being measured by the two sleep measures, it would be important to examine whether another objective sleep measure of WASO such as polysomnography shows a positive or negative association with next day processing speed performance. Despite the discrepant findings, however, given that insomnia is primarily diagnosed and monitored through self-reported sleep parameters, our results suggest that interventions specifically aimed at improving patients' perceptions of nighttime sleep maintenance may improve next-day cognitive performance in older adults with a history of chronic pain. Although sleep did not negatively affect SDMT performance in individuals without a history of chronic pain, our daily findings for this group are consistent with our previous results showing that poor sleep (ie, greater total wake time) is associated with better average processing speed and selective attention performance over 2 weeks of baseline.10 Specifically, in the current study, we found that greater self-reported WASO led to better next-day performance on SDMT. It is possible that increased WASO at night may lead to a hyperaroused state11 which then in turn leads to better next-day cognitive performance in those without a history of chronic pain, at least on a cognitive task that measures lower level processing.

Our prediction that mood would be negatively associated with cognitive performance, particularly for those with a history of chronic pain, was partially supported, but depended on the cognitive outcome that was measured. For instance, our results for the reasoning task suggest that overall mood rather than sleep may moderate cognitive performance in older adults with a history of chronic pain. These results agree with a previous study showing that depressive scores accounted for the greatest degree of variance in perceptions of cognitive difficulties in patients with chronic pain but extend these findings to objectively measured cognitive performance.30

Although we hypothesized that higher levels of pain severity would be associated with worse daily cognitive performance, particularly for participants with a history of chronic pain, this was generally not supported. In contrast to previous findings showing that pain severity was inversely associated with performance on a neuropsychological test of executive function in older adults with chronic pain conditions,13 we found that daily pain ratings did not affect daily reasoning performance. However, our results do agree with previous findings showing no association between pain severity and cognitive complaints in patients with chronic pain across the lifespan.30 Therefore, in older adults with insomnia, the emotional response associated with the experience of chronic pain may be a more important predictor of day-to-day higher-order cognitive function than pain severity. Given that the results reported here represent aggregate mood ratings over 2 weeks, it will be important for future research to examine the potential associations between daily mood ratings and cognitive performance, in order to determine the extent to which daily fluctuations in depressive symptoms affect day-to-day reasoning.

Results for daily verbal memory show different patterns of sleep and pain associations for individuals with and without a history of chronic pain. Given that participants without a history of chronic pain showed worse performance on verbal memory when they rated their pain as more severe suggests that fluctuations in pain perceptions may have a greater effect on memory when it is experienced acutely, rather than chronically. Further, it is possible that when pain is experienced over a prolonged period (ie, in chronic pain conditions), it no longer has the same daily effect on memory. In fact, in our participants with chronic pain, greater self-reported WASO led to better next-day performance on the verbal memory task. Therefore, increased arousal rather than pain perceptions may have a larger effect on verbal memory in these patients. Given that sleep efficiency was positively associated with next-day memory performance, and SOL showed no association with daily memory performance, total sleep time may be an important contributing factor to daily cognition in participants with a history of chronic pain. That is, although WASO may increase arousal, this should be considered in the context of amount of time spent sleeping. A higher sleep to wake ratio may be more important for patients with chronic pain relative to those without a history of chronic pain, in terms of the positive effect on next-day verbal memory performance.

For individuals without a history of chronic pain, our results suggest that morning pain severity may affect processing speed and selective attention as well as verbal memory. Our observation of daily greater morning pain being associated with worse performance on tasks measuring these abilities was expected, suggesting that interventions targeting pain rather than sleep in this group may improve daily processing speed/selective attention and memory performance. Because daily pain ratings did not influence similar cognitive performance in those with history of chronic pain, it is possible that sustained pain, such as that experienced by older adults with chronic pain, leads to a change in the association between pain and cognitive processing.

One potential limitation of the current study and the interpretation of results is the high level of education of the participant groups. Therefore, the reported results may not necessarily be representative of typical older adults with insomnia. Additionally, although our results did not support the hypothesis that higher pain ratings would lead to worse cognitive performance, particularly in those with a history of chronic pain, it is important to note that the mean ratings of pain were generally considered low for both participant groups (see Table 1). Therefore, it will be important for future research to examine whether associations between daily pain and cognition for those with a history of chronic pain are present in individuals with moderate to severe self-reported pain. Further, the goal of future work may be to directly compare the daily associations between sleep, mood, and cognition in those with moderate to severe pain relative to those with no pain or minimal pain. A final potential limitation concerns our sample size which was relatively small for each group. Thus, it will be important to compare the current findings with those in larger samples. However, sample sizes of approximately 30 or more have minimal effect on the accuracy of the standard error of fixed effects in multilevel models.31 This, combined with our use of daily data (14 days of observations for each participant) minimizes the effects of this potential limitation.

Our examination of daily associations between sleep, pain, and cognition is novel and suggests that history of chronic pain should be considered in the management of insomnia symptoms and secondary effects such as cognitive disturbances. In older adults with insomnia and a history of chronic pain, improving nighttime sleep may only benefit lower level cognitive performance, whereas reducing depressive symptoms may improve higher-level cognitive performance. Discrepancies in sleep parameters support the importance of assessing both objective and subjective sleep outcomes in the investigation of the effect of insomnia symptoms on cognition.

DISCLOSURE STATEMENT

Work for this study was performed at the University of Florida. All authors have seen and approved this manuscript. The project described was supported by Award Number AG024459 (Christina S. McCrae, PhD, PI) from the National Institute on Aging (NIA). The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIA. The authors report no conflicts of interest.

ABBREVIATIONS

BDI-II

Beck Depression Inventory, Second Edition

MLM

multilevel modeling

MMSE

Mini-Mental State Examination

SDMT

Symbol Digit Modalities Test

SOL

sleep onset latency

WASO

wake after sleep onset

REFERENCES

1 

Ohayon MM. Epidemiology of insomnia: what we know and what we still need to learn. Sleep Med Rev. 2002;6(2):97–111. [PubMed]

2 

Rustøen T, Wahl AK, Hanestad BR, Lerdal A, Paul S, Miaskowski C. Age and the experience of chronic pain: differences in health and quality of life among younger, middle-aged, and older adults. Clin J Pain. 2005;21(6):513–523. [PubMed]

3 

Taylor DJ, Mallory LJ, Lichstein KL, Durrence HH, Riedel BW, Bush AJ. Comorbidity of chronic insomnia with medical problems. Sleep. 2007;30(2):213–218. [PubMed]

4 

Dragioti E, Bernfort L, Larsson B, Gerdle B, Levin L. Association of insomnia severity with well‐being, quality of life and health care costs: A cross‐sectional study in older adults with chronic pain (PainS65+). Eur J Pain. 2018;22(2):414–425. [PubMed]

5 

Salthouse TA. When does age-related cognitive decline begin? Neurobiol Aging. 2009;30(4):507–514. [PubMed Central][PubMed]

6 

Haimov I. Association between memory impairment and insomnia among older adults. Eur J Ageing. 2006;3(2):107[PubMed Central][PubMed]

7 

Fortier-Brochu É, Beaulieu-Bonneau S, Ivers H, Morin CM. Insomnia and daytime cognitive performance: a meta-analysis. Sleep Med Rev. 2012;16(1):83–94. [PubMed]

8 

Riedel BW, Lichstein KL. Insomnia and daytime functioning. Sleep Med Rev. 2000;4(3):277–298. [PubMed]

9 

Nebes RD, Buysse DJ, Halligan EM, Houck PR, Monk TH. Self-reported sleep quality predicts poor cognitive performance in healthy older adults. J Gerontol B Psychol Sci Soc Sci. 2009;64(2):180–187. [PubMed]

10 

McCrae CS, Vatthauer KE, Dzierzewski JM, Marsiske M. Habitual sleep, reasoning, and processing speed in older adults with sleep complaints. Cognit Ther Res. 2012;36(2):156–164. [PubMed Central][PubMed]

11 

Bonnet MH, Arand DL. Hyperarousal and insomnia: state of the science. Sleep Med Rev. 2010;14(1):9–15. [PubMed]

12 

Buckalew N, Haut MW, Morrow L, Weiner D. Chronic pain is associated with brain volume loss in older adults: preliminary evidence. Pain Med. 2008;9(2):240–248. [PubMed]

13 

Weiner DK, Rudy TE, Morrow L, Slaboda J, Lieber S. The relationship between pain, neuropsychological performance, and physical function in community‐ dwelling older adults with chronic low back pain. Pain Med. 2006;7(1):60–70. [PubMed]

14 

Bair MJ, Robinson RL, Katon W, Kroenke K. Depression and pain comorbidity: a literature review. Arch Intern Med. 2003;163(20):2433–2445. [PubMed]

15 

Bierman E, Comijs H, Jonker C, Beekman A. Effects of anxiety versus depression on cognition in later life. Am J Geriatr Psychiatry. 2005;13(8):686–693. [PubMed]

16 

Beck AT, Steer RA, Garbin MG. Beck Depression Inventory-Second Edition. San Antonio, TX: The Psychological Corporation; 1996.

17 

Yesavage JA, Brink TL, Rose TL, et al. Development and validation of a geriatric depression screening scale: a preliminary report. J Psychiatr Res. 1983;17(1):37–49

18 

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. [PubMed]

19 

Murden RA, McRae TD, Kaner S, Bucknam ME. Mini-Mental State exam scores vary with education in blacks and whites. J Am Geriatr Soc. 1991;39(2):149–155. [PubMed]

20 

Lichstein KL, Riedel BW, Means MK. Psychological treatment of late-life insomnia. Annu Rev Gerontol Geriatr. 1998;18(1):74-110.

21 

McCrae CS, Curtis AF, Williams JM, et al. Efficacy of brief behavioral treatment for insomnia in older adults: examination of sleep, mood, and cognitive outcomes. Sleep Med. 2018;2(51):153–166

22 

Dworkin RH, Turk DC, Wyrwich KW, et al. Interpreting the clinical importance of treatment outcomes in chronic pain clinical trials: IMMPACT recommendations. J Pain. 2008;9(2):105–121. [PubMed]

23 

Smith A. Symbol digit modalities test (SDMT) manual (revised). Los Angeles, CA: Western Psychological Services; 1982.

24 

Spreen O, Strauss E. A Compendium of Neuropsychological Tests: Administration, Norms, and Commentary. Oxford University Press; 1998.

25 

Thurstone TG, Thurstone LL. Primary Mental Abilities Tests. Science Research Associates; 1962.

26 

Rey A. The clinical examination in psychology. Paris: Press Universitaire de France; 1964.

27 

Enders CK, Tofighi D. Centering predictor variables in cross-sectional multilevel models: A new look at an old issue. Psychol Methods. 2007;12(2):121–138. [PubMed]

28 

Bates D, Kliegl R, Vasishth S, Baayen R. Parsimonious mixed models. Retrieved from arxiv.org/abs/1506.049672015.

29 

Williams JM, Kay DB, Rowe M, McCrae CS. Sleep discrepancy, sleep complaint, and poor sleep among older adults. J Gerontol B Psychol Sci Soc Sci. 2013;68(5):712–720. [PubMed Central][PubMed]

30 

McCracken LM, Iverson GL. Predicting complaints of impaired cognitive functioning in patients with chronic pain. J Pain Symptom Manage. 2001;21(5):392–396. [PubMed]

31 

Scherbaum CA, Ferreter JM. Estimating statistical power and required sample sizes for organizational research using multilevel modeling. Organizational Research Methods. 2009;12(2):347–367