Association Between Socioeconomic Status and Comorbidities Among Patients With Rheumatoid Arthritis

Results of a Nationwide Cross-Sectional Survey

Anna Shin; Seunghwan Shin; Ji Hyoun Kim; You-Jung Ha; Yun Jong Lee; Yeong Wook Song; Eun Ha Kang


Rheumatology. 2019;58(9):1617-1622. 

In This Article


Patient Characteristics

Among 56 039 participants aged ≥19 years from 2007–2015 KNHANES, 4137 participants with missing information for RA were excluded. We identified a total of 1088 RA patients, which corresponds to the weighted number of 612 303 patients. The mean age was 58.0 years and 74.9% were female. Supplementary Table S1, available at Rheumatology online, shows the weighted prevalence of individual SES levels and comorbidities among RA patients. Overall, 53.7% belonged to the low-income group, 61.7% to the low-education group, and 77.5% to the urban dwellers.

When stratified by SES levels, patients of low vs high SES were much older and more comorbid (Supplementary Table S1, available at Rheumatology online). Due to substantial age difference between the two SES groups, we stratified the patients into ≥ and < median age (63 years). The weighted comorbidity distribution is shown for those aged <63 years (Table 1) and ≥63 years (Supplementary Table S3, available at Rheumatology online).

Comorbidity Distribution According to SES Levels in Those Aged <63 Years

In those aged <63 years (n = 531, weighted n = 371 157), ~70% were female and the mean age was 48.8 years. Due to residual age discrepancy within this age stratum (Table 1), a logistic regression model was used, further accounting for age effect, comparing the low vs high SES groups.

The most common comorbidities identified by the health interview were mental (high stress 35.1%, depressive mood 18.6% and suicide ideation 19.5%), cardiometabolic (hypertension 17.6% and dyslipidaemia 13.6%) and musculoskeletal (osteoporosis 10.6% and osteoarthritis 11.0%). Obesity (31.8%), hypertension (26.0%), hypercholesterolemia (15.2%) and hypertriglyceridemia (16.9%) were also commonly found by the medical examination (Table 1).

The age-adjusted OR (95% CI) associated with low vs high income was 3.09 (1.31, 7.29) for diabetes, 2.12 (1.13, 3.99) for osteoarthritis, 2.13 (1.01, 4.53) for depression, 2.68 (1.54, 4.65) for depressive mood and 3.01 (1.79–5.07) for suicide ideation based on the health interview. The age-adjusted OR (95% CI) was 2.04 (1.30, 3.20) for obesity, 2.30 (1.03, 5.16) for diabetes and 2.36 (1.28–4.34) for hypertriglyceridemia based on the medical examination (Table 2). Low education was associated with suicide ideation (OR 2.25, 95% CI 1.14, 4.44) and tended to be associated with osteoarthritis (OR 2.08, 95% CI 0.99, 4.40) after adjusting for age. We observed no association between area of residence and comorbidities before or after adjusting for age.

Comorbidity Distribution According to SES Levels in Those Aged ≥63 Years

In those aged ≥63 years (n = 557, weighted n = 241 145), 83.1% were female and mean age was 72.2 years (Supplementary Table S3, available at Rheumatology online). We found no significant difference in comorbidity distribution according to SES after adjusting for age and/or gender.

Health-related Behavior and Healthcare Utilization

To examine the association mechanism between comorbidities and SES among those aged <63 years, we compared health-related behaviors (exercise pattern, drinking or smoking habits) and healthcare utilization patterns between high and low SES levels (Supplementary Table S4, available at Rheumatology online).

A lower rate of vigorous exercise was associated with low income (P = 0.04) or tended to be associated with low education (P = 0.08). We observed no difference in the overall drinking frequencies (P = 0.23 for a trend test) or smoking habits (P = 0.27 for a trend test) according to income levels.

Failed access to necessary medical services was numerically more frequent in low SES groups (23.8% vs 18.1% in the low and high income group, respectively; 23.8% vs18.0% in the low and high education group, respectively). Outpatient clinic visits within two months from the interview tended to be more frequent in the low income than in the high income group (P = 0.07 for a trend test), potentially reflecting worse health conditions in the former group. There was no difference in the number of patients hospitalized during the past year.