Background/Purpose: Rheumatoid Arthritis (RA) is associated with excess mortality.  Indigenous North Americans (INA) in our region have high RA prevalence rates and young age at onset yet experience disparities in arthritis treatment.  We determined if comorbidity and mortality were increased in INA with RA and the factors associated with mortality in RA.

Methods: Using administrative health data from our region (years 2000-2010; population 1.1 million), a previously validated case definition for RA, and the INA identifier from Indian Affairs, we identified cohorts of incident (N=4195) and prevalent (N-8095) RA.  Comorbidity was determined using a modified Charlson Comorbidity Index (mCCI) categorized as 0, 1, 2, 3 or more, and the John Hopkins mental and physical Major Adjusted Disease Groups (mADG, pADG).  Regional income quintiles were used as a proxy for socioeconomic status.  Death, age at death and cause of death were identified in prevalent and incident RA; duration of RA at death was determined in incident RA.  Crude all-cause mortality rates were adjusted to age and sex or to age, sex and last visit mCCI.  Annual mortality rates between INA and nonINA and persons with or without RA were compared using Student T tests.  Cox proportional hazards models evaluated contributors to death in RA controlling for age, sex, ethnic group, income quintile, and comorbidity. Odds Ratio(OR) with 95% confidence limits (CL) are reported.  

Results: In spite of a young onset age (INA 42 vs nonINA 55 yr p<0.001), INA were more likely to have nonRA comorbidity (mCCI >0) than nonINA at baseline (39% vs 31% OR 1.43 CL 1.25-1.64 p<0.0001) but equally likely to have multiple comorbidities at last visit (both 22% OR 0.99 CL 0.84-1.16).  More INA than nonINA reported mADG at baseline (27% vs 19% OR 1.59 CL 1.37-1.84) and last visit (26.9% vs 22.8% OR 1.29 CL 1.12-1.49). pADG rates were similar at baseline (39% vs 42% OR 0.9 CL 0.79-1.02) and last visit (49% vs 52% OR 0.9 CL 0.79-1.02).  Between 2000-2010 1068 prevalent RA patients died (96 (9%) INA; 972 (14%) nonINA including 301 incident RA (23 (4%) INA; 278 (8%) nonINA).  Prevalent RA INA were much younger at death than nonINA (56 (CL 54-59)vs 77 (CL 76-77) years; p<0.0001), a trend seen for males and females each year even after adjusting for mCCI. Cause of death was similar for INA and nonINA with a trend to less cancer deaths and more “other” deaths in INA.  Age and sex (and mCCI) adjusted mortality rates decreased in the general population yet increased for persons with RA.  Age, sex and mCCI adjusted annual mortality rates were higher in INA than nonINA with RA.  In cox proportional hazards models, increasing comorbidity, both pADG (1.64 (1.29-2.08)) and mADG (1.56 (1.19-2.04)), older age, female sex and lower income predicted death, but INA ethnicity did not.  

Conclusion: Persons with RA, in particular INA, have increased mortality partly explained by increasing mental and physical comorbidity.  The high rate of comorbidity at an early age and young age at death in INA RA is striking.  The independent influence of mental comorbidity on mortality suggests a complex social-biologic phenomenon of particular relevance to INA given their unique social stressors that needs to be addressed as an early step to improving outcomes for this vulnerable population.

« Back to 2015 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/increased-mortality-in-indigenous-north-americans-persons-with-rheumatoid-arthritis-is-partially-explained-by-psychiatric-and-physical-comorbidity-a-population-based-study/