Interactions of social determinants
The authors analyzed retrospective data from 1,684 pediatric patients in the Childhood Arthritis and Rheumatology Research Alliance (CARRA) registry covering the period of April 2015 to February 2020. All study participants had been diagnosed with pJIA. Symptom onset occurred before age 16, and at least five joints were involved. The authors excluded patients who had been diagnosed with other systemic inflammatory or autoimmune diseases.
The authors defined exposure to a high level of community poverty as living in a ZIP code where at least 20% of residents lived at or below the federal poverty level. The authors also collected data on household income, although these data were missing for more than a quarter of participants (27%) and were therefore included only in sensitivity analyses. They used the clinical Juvenile Arthritis Disease Activity Score–10 (cJADAS-10) and the Child Health Assessment Questionnaire (CHAQ) to assess disease activity and disability at baseline and 6 and 12 months later. A cutoff of 2.5 on the cJADAS-10 distinguished mild disease activity from moderate to high disease activity, and a CHAQ score of 0.25 was the cutoff for having functional disability.
Among those who reported household income, just over half the cohort had an income of at least $50,000. The study population was 74% White, and more non-White patients lived in high-poverty communities (36.4%) than did White patients (21.3%). Patients whose families had no more than a high school education (23.1% vs. 13.7%) and those with public insurance (43.0% vs. 21.5%) were also over-represented in poorer communities.
The median cJADAS-10 scores declined overall during patients’ first year of therapy. However, those with public insurance, a lower family education level, or residency in poorer communities made up the greatest proportion of patients who continued to have moderate to severe disease activity a year after diagnosis.
The unadjusted calculations showed that children living in high community poverty had 1.8 times greater odds of functional disability (odds ratio, 1.82; P < .001). However, after adjustment for age, sex, race/ethnicity, insurance status, family education, rheumatoid factor, and cyclic citrullinated peptide antibody, the association lost statistical significance (P = .3). Community poverty level was not associated with disease activity before or after adjustment.
“Race was adjusted for as a confounder; however, the association between race/ethnicity and social determinants of health is likely more complex,” Dr. Feldman said. “Interactions, for example, between individual race and area-level poverty could be investigated.”
Odds of persistent function disability were 1.5 times greater for children with public insurance (adjusted OR, 1.56; P = .023) and 1.9 times greater for those whose families had a lower education level (aOR, 1.89; P = .013). Children whose race/ethnicity was indicated as being other than White had more than double the odds of higher disease activity (aOR, 2.48; P = .002) and were nearly twice as likely to have persistent functional disability (aOR, 1.91; P = .031).
Future directions
Dr. Soulsby was struck by the difference in statistical significance between individual-level poverty, as measured by household income, and community-level poverty. “It’s interesting because it may suggest that both of these forms of poverty are different and have different impacts on disease,” he said. Dr. Balmuri elaborated on the nuances and interactions that exist with social determinants of health and how objective outcomes, such as disease activity as measured by clinical tools, can differ from subjective outcomes, such as patients’ reports of pain, daily disability, and social experiences.
“The human condition is far more complicated, unfortunately, than any dataset could have on their own collected,” Dr. Balmuri said. She said she plans to expand her pJIA research into other social determinants of health. “It’s first about getting people’s eyes and minds open to something we see every day that, for some reason, sometimes people are blinded to, [using] the data that we do have, and then our hope is to build upon that.”
Dr. Feldman noted that ZIP codes, which were used as a proxy for community poverty, may not provide the best perspective regarding a patient’s neighborhood, because significant variation may exist within a single ZIP code, which is something the authors noted as well. The investigators were limited in the data available from the registry, and Dr. Balmuri and Dr. Soulsby suggested that 9-digit ZIP codes or census tracts might better capture neighborhood deprivation.
The research was funded by the Arthritis Foundation and the National Institute of Arthritis and Musculoskeletal and Skin Diseases. Dr. Feldman has received research support from Pfizer and the Bristol-Myers Squibb Foundation. Dr. Soulsby and Dr. Balmuri have disclosed no relevant financial relationships.
A version of this article first appeared on Medscape.com.