Table 2.
Observational studies using longitudinal data to assess the relation of chronic pain and chronic pain conditions to subsequent cognitive decline and incident cognitive impairment and dementia: characteristics and key findings
| First Author, y, Country | Design and Follow-up Period (y) | Data source | Study sample size | Population characteristics | Chronic pain measure | Outcome | Main Results AOR/ARR/AHR (95% CI) | Factors adjusted for | ||
|---|---|---|---|---|---|---|---|---|---|---|
| Osteoarthritis (OA) | ||||||||||
| Innes 2020, USA [94] | Retrospective cohort (elders enrolled in fee-for-service (FFS) Medicare, 11 pooled cohorts). f/u: 2 y | Linked Medicare claims and Medicare Current Beneficiary Survey files (2001–2015) | 16,934 beneficiaries (1,149 ADRD, 4,545 OA) | Non-institutionalized adults 65 + y, ADRD-free at baseline, continuously enrolled in FFS Medicare | OA diagnosed at baseline (ICD-9-CM claims:≥2 out-patient claims≥ 90 days apart or≥ 1 inpatient claim); diagnosed joint, neck/ back, or neuropathic pain (ICD-9-CM) | Incident ADRD; Diagnosed y 2–3 (ICD-9-CM 290.0–290.3, 331.0–331.2, 331.7,331.8 or affirmative response to Health Status question re AD¥. | AORs for ADRD: OA: 1.23 (1.06,1.42) versus No OA or pain: OA+ pain: 1.31 (1.08,1.58) OA, no pain: 1.23 (0.94,1.66) No OA+pain: 1.13 (0.96,1.33) Incl depression, anxiety OA: 1.14 (0.98,1.32) versus No OA or pain: OA+pain: 1.20 (0.99, 1.45) Competing risk of death models: Similar AORs |
Age, sex, race/ethnicity, education; marital status, insurance, region, smoking, BMI, medications, diabetes, hypertension, heart failure, IHD, hypertension, kidney disease, cancer, diabetes, Parkinson’s disease, headache, migraine COPD, RA, SLE, hx of stroke, TBI. Also incl depression, anxiety, sleep disorders | ||
| Wang 2018, Taiwan [86] | Nested case control (random sample of 2 million pp; controls matched on age, sex). f/u: up to 10 y | National Health Insurance Research Database (NHIRD) (2001–11) | 7,854:2,618 AD, 5,236 age and sex-matched controls (2,678 OA) | Mean age 76.1 y; 59% F | Pain conditions (dx ≥ 1 y before AD): OA (ICD-9-CM 714, 715, 720, 721), OA/osteoporosis (OS) cluster (ICD-9-CM) | Incident AD (ICD-9-CM 331.0, 290.0, 290.2, 290.3, on AD meds approved for reimbursement) | AORs for AD: OA: 1.09 (0.95,1.26) OA/OS: 1.23 (1.07,1.41) | Age, sex | ||
| Chen 2018, Taiwan [100] | Matched case control (Matched age and sex). f/u: up to 10 y | Longitudinal Health Insurance Database (LHID; data drawn from NHIRD) (2000–10) | 71,260 adults (10,180 dementia, 61,080 controls matched on age, sex, index date): 37,051 OA | ≥40 y | Rheumatic Disorders, including OA: ICD-9-CM | All-cause dementia (2001–10) (ICD-9-CM codes 290.0–290.4, 294.1, 331.0) | AORs for dementia: OA: 1.47 (1.40–1.54) | Age, sex, comorbidities (not specified) | ||
| Huang 2015, Taiwan [87] | Retrospective matched cohort; controls matched 2:1 on age, sex. f/u: up to 4 y | National NHIRD-LHID (2004–2011) | Total: 105,447 OA (2004–7): 35,149 Controls: 70,298; 1,070 incident dementia | ≥ 18 mean age not reported | Osteoarthritis (5 +OA diagnoses (ICD-9) or hospitalization with OA primary dx) | All-cause dementia after OA dx (≥2 dx and 1 primary dx in hospital) | AHR (incident dementia): OA: 1.25 (1.10–1.43) | Baseline age, sex, urbanization; diabetes, dyslipidemia, RA, SLE, COPD, stroke (CVA), hypertension, IHD, Parkinson’s disease. | ||
| Fibromyalgia (FMS) | ||||||||||
| Tzeng 2018, Taiwan [88] | Retrospective matched cohort (controls matched on age, sex). f/u: Up to 10 y | NHIRD-LHID (2000–2010) | Total: 166,448 (41,614 Fibromyalgia (FMS); 124,836 controls). Exclude pts with ADRD before FMS dx; or with SLE, RA, SS. 6,123 with incident dementia. | ≥50 y (mean age not reported) 87.6% F | Newly diagnosed Fibromyalgia (ICD-9-CM ICD-9-CM 411.1, 413, 414.0, 414.8–414.9 + 3+outpt visits) in 2000 | All-cause dementia (2000–10); AD; VaD; non-VaD (ICD-9-CM 290.0, 290.10, 290.11, 290.12, 290.13, 290.20, 290.21, 290.3, 290.41, 290.42, 290.43, 290.8, 290.9, 331.0) | AHRs: Dementia: 2.77 (2.61–2.95); AD: 3.35 (2.57,4.32); VaD: 3.14 (2.60–3.78) non-VaD: 2.72 (2.55,2.90) | Age, sex, urbanization, income, region of residence insurance premium; diabetes, dyslipidemia, COPD, CKD, hypertension, IHD, MI, PVD, AF, HF, stroke, depression, anxiety & other MH disorders; head injury; Parkinson’s disease; cancer, inflammatory, connective tissue, rheumatologic, liver, kidney, GI, thyroid disease; Anti-DM drugs, antihypertensive drugs, statins | ||
| Headaches | ||||||||||
| Morton 2019, Canada[89] | Retrospective cohort: Of 1,790 pts (60.5% of those contacted), 1,355 w/o CI = >961 (71%) w/complete data = >245 died/LtF, 37 CI (non-ADRD) = >679 final sample. f/u: 1–5 y | Manitoba Study of Health and Aging | 679 adults cognitively intact at baseline with complete data on covariates, migraine | adults 65 + y at baseline | Migraine (Self-reported history of at baseline) (N= 72) | Dementia (DSM = IV) (N= 51) AD (NINCDS-ADRDA) (N= 34) VaD (NINDS-AIREN) (N= 12) | AORs for: Dementia: 2.97 (1.25, 6.61); AD: 4.22 (1.59,10.42); VaD: 1.52 (0.20,7.23) Adjusted for age, education (AD), & stroke (dementia), depression (VaD) | (See Results) Self-reported data on: age, sex, education, depression (‘confounding’); hypertension, diabetes, stroke, myocardial infarction and other heart conditions (‘intervening’) | ||
| Tzeng 2017, Taiwan [90] | Retrospective matched cohort (controls matched on age, sex). f/u: Up to 10 y | National level NHIRD-LHID (2000–2010) | Total: 14,480 (3,620 with newly diagnosed headaches (2000); 10,860 controls; 603 with incident dementia | ≥ 20 y; mean age not reported | Primary headache including Migraines (M) and Tension-type headache (TTH) | All-cause incident dementia; AD; VaD; non-VaD | Competing risk model: AHRs for: Dementia Any = 2.05 (1.71,2.46); M: 2.00 (1.57,2.53) TTH: 1.77 (1.41,2.22) Non-VaD: Any: 2.11 (1.75,2.55) M: 2.06 (1.53,2.51) TTH: 1.88 (1.49,2.37) VaD (N= 46 tot): NS | Age, sex, urbanization, location, DM, dyslipidemia, COPD, hypertension, IHD, AF, HF, MI, PAD, stroke, depression and other MH disorders, head injury, Parkinson’s disease, cancer, rheumatic, kidney, liver, and specific inflammatory/ connective tissue disease, epilepsy, pain, gout. | ||
| Yang 2016, Taiwan [91] | Retrospective matched cohort (controls frequency-matched 4:1 on age, sex, TTH dx date). f/u: Up to 10 years (until ADRD dx, death, or EOS), avg 8.1 y | National level NHIRD-LHID (2000–2010) | Total: 69,540 (13,908 with newly diagnosed TTH (2000–6); 55,632 TTH-free controls; 2,237 incident dementia | ≥ 20 y; mean age not reported; | Tension-type headaches (TTH) (ICD-9-CM 307.81 and 339.1) diagnosed 2000–6 (index date) | All-cause incident dementia (ICD-9-CM 290, 294.1, and 331.0); AD; VaD; non-VaD | Competing risk model: AHRs for: ADRD: 1.15 (1.05,1.27) Women: 1.25 (1.11,1.42)>65 y: 1.13 (1.01,1.27); non-VaD: 1.21 (1.09.1.34) AD, VaD: NS | Age, sex, diabetes, dyslipidemia, COPD, hypertension, IHD, AF, HF, stroke, depression, head injury, Parkinson’s disease and migraine | ||
| Straete Rottereng 2015 Norway [92] | Retrospective cohort. Of 64787 HUNT 2 pts (70% of those invited), 52,230 (56%) complete head-ache Q.¥¥ = >9617 (18%) died/LtF = >26197/52541 (62%) complete HUNT 3 Qs = >15697 55–89 y. f/u: up to 15 y | Nord-Trondelag Health Surveys: 1995–7 (HUNT2) and 2006–8 (HUNT3) and Dementia Regist (1998–2011); all residents≥20 y of NordTrøndelag invited | Reference group: 15,601 adults 55–89 y with headache info in HUNT2 and participate in HUNT3; 746 HUNT 2 pts in Dementia Regis; 96 confirmed healthy, no ADRD (HUNT3) | Age 55–89 y | At baseline (HUNT2): Any headache; migraine; non-migraine headache | Dementia (Dementia registry); confirmed non-demented (HUNT3) | AORs for any headache: Dementia: 1.24 (1.04–1.49) No Dementia: 0.62 (0.39,0.98) | Age, sex, education, HADS score, smoking, severe comorbid condition (Final analyses); also consider BMI, alcohol use, BP, physical activity. | ||
| Hagen 2014, Norway [95] | Prospective cohort. Of 64787 HUNT 2 pts (70% of those invited), 52,222 (56%) complete head-ache Q¥¥ =>51383 w/o CI or ADRD. f/u: up to 15 y (avg to dementia dx = 8.6 y) | Nord-Trondelag Health Surveys: 1995–7 (HUNT2) and Dementia Registry (1997–2010); all residents≥20 y of NordTrøndelag invited | 51,383 adults responding to headache questionnaire; 378 with incident dementia (63 VaD, 52 mixed dementia (D), 180 AD). Identified in Dementia Regis | Age≥20 y | At baseline (HUNT2): Any reported headache; migraine; non-migraine headache within past 12 months | Dementia-all cause, VaD, AD, dementia with Lewy bodies; FT D. Conservative dx (MRI, clin expert; ICD-10; NINCDS-ADRDA criteria required for AD, NINDS-AIREN for VaD) | AHRs for headache at baseline: All dementia: 1.3 (1.1–1.7) VaD: Any: 2.3 (1.4–3.8); M: 2.9 (1.3–6.6); Non-M:2.1 (1.2–3.7) Mixed D:2.0 (1.1,3.5) AD, other dementia subtypes: No association | Final models incl: All dementia: Age, sex Subtypes: age, sex, education, smoking, total HADS score Also info on: marital status, physical activity, alcohol use, anxiety, depressive sx; BMI, BP, lipids, glucose; HT meds; self-reported MI, DM, angina, stroke; daily med use. | ||
| Chuang 2013, Taiwan [93] | Retrospective matched cohort (adults with incident migraine frequency-matched to those without migraine on age, sex). f/u: up to 12 y until EOS, death, loss to f/u, or (controls) new migraine dx | NHIRD (1998–2010) | 167,340 adults: 33,468 with incident migraine; 133,872 with- out migraine. 689 with incident dementia. | Avg age = 42.4 (Migraine) and 42,1 (non-migraine); 71.3% F. All dementia-free at baseline | Newly diagnosed migraine (ICD-9 346) 1998–2010. | All cause dementia (ICD-9 290, 294.1 and 331.0) | AHRs for migraine: ADRD: 1.33 (1.22,1.46) | Age, sex; baseline diabetes, hypertension, coronary artery disease, head injury and depression | ||
| Chronic pain, unspecified/site-specific | ||||||||||
| Yamada 2019, Japan [96] | Prospective cohort (2013 baseline). f/u: 3 y to EOS (2017), leave area where registered, dementia dx, or death | Survey (mailed self-administered questionnaires) study of Japanese residents≥65 from 30 local govts (71% RR) in 2013 (baseline), linked to long- term care insurance registry (JAGE) | 14,627 dementia-free Japanese residents≥65 who applied to LT care but did not receive benefits (482 with incident dementia). | 65 + y at baseline, excl those with hx of stroke, cancer, injuries, depression, PD, dementia; who need daily living support or lack ADL info | Self-reported knee pain, low back pain, defined as presence of knee pain/low back pain in last year that interfered with daily activities (2 questions). | Incident dementia: data from LT care insurance; dx based on standardized home assessment (ADL, instrumental ADL; cognitive function, mental/ behavioral disorders** | Dementia: AHRs for Knee pain: 1.32 (1.06–1.64); appear strongest in those 65–79 (AHR 1.73 (1.11,2.68) and those who do not walk regularly (although interactions for latter NS) Back Pain: 0.79 (0.63–0.99) appear strongest in > 80 (AHR = 0.5 (0.3–0.8)) | Self-reported age group, sex, education, marital status, income, employment, loss events, social interaction, BMI, alcohol use (Y/N/ex-drinker), smoking ((Y/N/ex), DM, hyper-tension; mood/anxiety disorder (Kessler Psych Distress Scale (K6) 13+). Regular walking (modifier) | ||
| Veronese 2018, England [101] | Longitudinal cohort. 9,432 (85% of original cohort) complete WAVE 2, 8,960 with complete data = >2,429 died/LtF, 16 ADRD≠≠ = >6615 (60% original cohort) in WAVE 4. f/u: 4 y in 2 data waves (Wave 2:2002–3; Wave 4:2008–9) | English Longitudinal Study of Ageing (ELSA)-nationally representative ongoing cohort study; nurse visit+in person interviews | Total: 6,515 community dwelling adults | ≥50 y, avg 65 y; 57.3% women | Resp to question: “often troubled by pain?” If Y, what intensity (mild, mod, sev) | Change in cognitive function (verbal fluency, memory (immediate/ delayed recall; processing speed (letter cancellation test) | Pain (Y/N) and cognitive decline: No association overall Severe pain and memory decline: −0.36 (−0.68; −0.04), p = 0.04 | Baseline: Age, sex, marital status, education, household wealth, race, smoking, physical activity, alcohol, disability, BMI, diabetes, lung disease, asthma, hypertension, IHD, AF, HF, stroke, depression (CES-D), Parkinson’s disease, cancer, arthritis, osteoporosis (self-report except BMI)- Diffs in all but 2 vars (PD, cancer) | ||
| van der Leeuw 2018, USA [97] | Prospective cohort; of 590 CCMA pts enrolled 2011–17, 521 with baseline data on pain, cognition = >441 w/o MCI, MCR, or ADRD.≠≠≠ f/u: 1–6.2 y, mean = 2.75 y | Central Control of Mobility in Aging (CCMA) (lower Westchester Co): Community-dwelling adults | Total: 441; 285 with pain at baseline; 56 with incident major cognitive impairment (CI) | ≥65; avg 76 y; free of ADRD, MCI, or MCR at baseline. | MOS SF-36 pain severity scale, analyzed as continuous var and in tertiles/ quartiles (low, mild, moderate, severe) | Incident major cognitive impairment (CI) (Repeatable Battery for the Assessment of Neuropsychological Status (RBANS); TMT Delta) | Major CI: No association with presence of pain (Y/N); In those with pain: Severe Pain: AHR= 3.47 (1.42–8.46) (versus low pain) for major memory impairment. | Age, sex, race, medications (incl analgesics, neuropathic/psych medications, sleep meds, antidepressants), General Health Score (summed conditions: DM, dyslipidemia, COPD, hypertension, IHD, AF, HF, stroke, PD, cancer, RA, etc.), depressive sx (GDS) | ||
| Ezzati 2018, USA [98] | Retrospective cohort Include EAS pts enrolled 1994–2015 with complete info on pain intensity/ interference and≥1 annual f/u. (cohort RR, LtF/deaths, missing data rates NR). f/u: 1–16.5 y; mean = 4.4 y | Einstein Aging Study (EAS), cohort of community-dwelling adults≥70 y in Bronx County, NY; dementia-free at baseline | Total: 1,114; 114 with incident dementia; 98 with probable AD | ≥70 y, avg 78 y | Pain intensity score (1–6); Pain interference (1–5) in past 4 weeks (2 q from SF-36): 91.4% with arthritis versus 60% in pain versus comparison group. | All-cause dementia; AD Assessed using DSM-IV criteria; also require meeting NINCDS-ADRDA criteria | Pain intensity: No association with AD/ADRD Pain interference: AHRs ADRD: 1.36 (1.07–1.73); Also incl depression, NSAIDs: 1.32 (1.03,1.69); <3 y versus≥3 y: 1.28 (0.69–1.83) versus 1.55 (1.11–2.2) AD: 1.26 (1.04–1.54) | Age, sex, race/ethnicity, education; medical comorbidity index (0–9): hypertension, diabetes, stroke, MI, angina, congestive heart failure, Parkinson’s disease, RA, COPD) | ||
| Whitlock 2017, USA [99] | Retrospective cohort: Of 12,058≥62 y (2000 wave), 10,065 with (non-proxy) data on pain, cognition (1998,2000 waves). f/u: up to 12 y until death/drop-out/EOS (2000–2012) | Health Retirement Survey (HRS, 2000–12): Population-based sample of community-dwelling older U.S. residents. In person/ telephone interviews ca. every 2 y | Total: 10,065; 1,120 with persistent pain (10.9 weighted %) | ≥62 y (2000 Wave); avg 73 y (interquartile range 67–78 y); 60% F | Persistent pain: troubled by mod/sev pain in 1998 and 2000 interviews (from items: “Are you often troubled with pain?” and “How bad is the pain most of the time: mild, moderate or severe?”) | Composite memory score and Dementia probability scores: developed using cognitive tests & interviews, i.e., using core HRS questions and models derived from ADAMS cohort. | Persistent pain: Associated with mean: 9.2% (CI 2.8–15%) faster cognitive decline; 7.7% (0.55–14.2%) faster increase in dementia probability after 10 y, corresponding to absolute 2.2% higher risk. | Baseline age, sex, race/ethnicity, education, marital status; financial assets; smoking, alcohol; med comorbidities (self- reported physician dx of DM, hypertension, COPD, heart disease, stroke, cancer [excl minor skin cancers]); depressive sx (CES-D), limitations in ADL. | ||
AD, Alzheimer’s disease; ADL, Activities of daily living; ADAMS, Aging, Demographics, and Memory Study; ADRD, Alzheimer’s disease and related dementia; ADRDA, National Institute of Neurological and Communicative Diseases and Stroke/Alzheimer’s Disease and Related Disorders Association; AF, atrial fibrillation; AHR, adjusted hazards ratio; AOR, adjusted odds ratio; ARR, adjusted relative risk; avg, average; CI, confidence interval; COPD, chronic obstructive pulmonary disease; DM, diabetes; dx, diagnosis; EOS, end of study; excl, excluding; F, female; f/u, follow-up; HF, heart failure; hx, history; HT, hypertensive; IHD, ischemic heart disease; incl, including; LtF, lost to follow-up; MCR, motoric cognitive risk syndrome; mod/sev, moderate/severe; MI, myocardial infarction; NINCDS-NINDS-AIREN, Neurological Disorders and Stroke (formerly NINCDS) and Association Internationale pour la Recherché et l’Enseignement en Neurosciences; NR, not reported; OA, osteoarthritis; RA, rheumatoid arthritis; RR, response rate; SLE, systemic lupus erythematosus; SS, Sjogren’s syndrome; sx, symptoms; VaD, vascular dementia; w/o, without; y, years;
*
Virtually universal;
**
Reported to correlate well with independent. physician panel assessment.
¥
’Has a doctor ever told you that you have Alzheimer’s?’.
¥¥
Those responding to headache questionnaire were younger, more likely to be women, and had higher socioeconomic status versus non-responders.
≠≠
Those excluded at wave 4 due to missing data or death were significantly older, were more likely to report pain, and scored significantly worse in all cognitive tests at wave 2.
≠≠≠
Eligible participants were significantly younger and more educated versus those excluded.