Cognitive Impairments in Immune-Mediated Inflammatory Diseases

Abstract

Objective: Studies suggest an increased prevalence of cognitive impairment (CI) among people with rheumatoid arthritis (RA) and other immune-mediated inflammatory diseases (IMIDs). Prior studies are limited by convenience-sampling and a lack of control for confounding variables of age, education, gender, race, and the genetic risk factor APOE-e4. This thesis examines CI in RA and IMIDs addressing these limitations in three stand-alone but related papers. Methods: Chapter one examines the validity of self-reported RA using Medicare Parts A & B claims as a referent standard and three algorithms to identify RA using international classification of diseases, clinical modification codes (ICD-9-CM). Additional self-report questions about care received for arthritis are used to try to improve the validity of self-reported RA. Chapter two examines the prevalence of of CI among people wth RA using a cross-sectional analysis at the 2016 survey wave. Two ICD-9-CM or ICD-10 codes identify RA in the preceding two years. The Langa-Weir Classification is used to classify cognitive status as normal, cognitively impaired non-dementia (CIND), or dementia based on a brief neuropsychological battery for self-respondents and informant reports for proxy respondents. The odds of CI in RA compared to older adults without RA are examined using unadjusted and adjusted logistic regression, controlling for age, education, gender, and race. Chapter three examines the risk of Alzheimer’s disease (AD) using time-to-event analysis in the IMIDs rheumatoid arthritis, ankylosing spondylitis (AS), psoriatic arthritis (PSA), Crohn’s disease, and ulcerative colitis (UC). Respondents with IMIDs were identified between 2006-2009 using two ICD-9-CM codes. AD was identified using the Chronic Conditions Warehouse criteria and time of first diagnosis. AD-free cohorts of IMID and non-IMID respondents were followed from 2009-2014 for incident AD, and were censored at death, the last day of their Medicare records, or the study conclusion. The risk of AD in IMID was calculated using unadjusted and adjusted Cox proportional hazards models, adjusted for age, gender, education, race, and APOE-e4. Results: In paper one, 345 of 3768 eligible respondents self-reported a physician diagnosis of RA. Across all three algorithms for identifying RA cases from Medicare records, the PPV of self-report ranged from .05 to .16., sensitivity from .23 to .55., and kappa statistic from .07 to .15. Additional self-reports regarding arthritis care improved validity measures; however, the values remained low. For paper two, Medicare records identified a 3.36% prevalence of RA (150 of 4,462 eligible respondents). While age, gender, education, and race independently predicted CI status, we found no difference in CI prevalence by RA status (prevalent CI in RA = 36.7% vs. 34.0% in non-RA); adjusted OR=1.08, 95% CI 0.74-1.59, p=.69). In paper three, 171 (6.02%) of 2,842 eligible respondents were classified with IMIDs. Over 6 years, 9.36% of IMID respondents developed AD compared to 8.57% of controls (unadjusted hazard ratio (HR): 1.09, 95% CI .66-1.81, p=0.74). Adjusted HR 1.27 (95% CI 0.76-2.12, p=0.35). Age (HR for ten-year increment 3.56, p<.001), less than high school education (HR 1.70, p=.007), and APOE-e4 (HR 2.61, p<.001 for one or two copies), were statistically significant predictors of AD. Conclusion: Most older adults who self-report RA do not have a Medicare claims history consistent with that diagnosis. There was no evidence of an association between RA and CI, and respondents with IMIDs did not have an increased risk of developing Alzheimer's disease over a 6-year period.