Comorbidity in arthritis.

OBJECTIVE

To describe the relative frequency of selected comorbidities in 2 population based prevalence cohorts of patients with rheumatoid arthritis (RA) and osteoarthritis (OA) compared to age and sex matched community controls.

METHODS

Using the population based data resources of the Rochester Epidemiology Project, we assembled 3 prevalence cohorts of all residents of Olmsted County, Minnesota, with RA (1987 American College of Rheumatology criteria) and age and sex matched controls without arthritis on January 1, 1965, January 1, 1975, and January 1, 1985. Cases and controls were followed longitudinally through their complete (inpatient and outpatient) medical records beginning 10 years prior to the prevalence (or index) date until death, migration from the county, or January 1, 1995. Comorbidity was assessed yearly using the Charlson Comorbidity Index and the Index of Co-existent Diseases (ICED). Descriptive statistics were used to illustrate the baseline characteristics of the study population and the frequency of individual comorbidities in each of the 3 groups over the followup period. Cox proportional hazards modeling was used to assess the risk for each individual comorbidity among patients with arthritis compared to controls and to identify significant predictors of an increase in comorbidity level over time.

RESULTS

Our study population included 450 RA, 441 OA, and 891 control subjects. The age and sex distributions of cases and their controls were similar. Over the followup period, patients with RA had a higher likelihood of developing congestive heart failure, chronic pulmonary disease, dementia, and peptic ulcer disease, while cases with OA had a significantly higher risk of developing peptic ulcer disease and renal disease. Among patients with either RA or OA, age, male sex, and baseline comorbidity were significant predictors of a rise in comorbidity. The presence of RA was a highly significant predictor of a rise in comorbidity from one year to the next, even after controlling for the effects of age, sex, and baseline comorbidity (p = 0.0004 for the Charlson and p = 0.006 for the ICED).

CONCLUSION

These data indicate that the burden of illness among people with arthritis is higher than for nonarthritics and that this burden appears to be increasing over time, particularly in RA. These results suggest that specialized chronic disease care will be increasingly important for the future health care needs of people with RA.