Association of Genetic Risk for Rheumatoid Arthritis With Cognitive and Psychiatric Phenotypes Across Childhood and Adolescence

Abstract

Importance: The association of rheumatoid arthritis (RA) with cognitive and psychiatric phenotypes has been recognized. However, it is not known whether these phenotypes are a consequence of disease-related factors, such as pain, or reflect shared etiological factors.

Objective: To investigate whether genomic risk for RA is associated with cognitive and psychiatric symptoms in children and adolescents.

Design, setting, and participants: This cohort study analyzed data from 3296 to 5936 adolescents (depending on outcome) from the Avon Longitudinal Study of Parents and Children. Clinical and questionnaire data were collected periodically from September 6, 1990, with collection ongoing, and analyzed from August 21, 2017, to May 21, 2018.

Exposures: Polygenic risk scores (PRSs) for RA.

Main outcomes and measures: Measures of cognition (including IQ, working memory, verbal learning, processing speed, problem solving, selective attention, and attentional control) and psychopathology (including anxiety, depression, negative symptoms, psychotic experiences, attention-deficit/hyperactivity disorder, and hyperactive and inattentive symptoms) in childhood and adolescence.

Results: Polygenic risk scores for RA were generated for 7977 children and adolescents (3885 [48.7%] female). Of these 7977 participants, 9 (0.11%) had a known diagnosis of RA at age 22 years. Increased PRS for RA was associated with lower total IQ (β, -0.05; 95% CI, -0.07 to -0.02; P < .001), performance IQ (β, -0.03; 95% CI, -0.06 to -0.005; P = .02), and verbal IQ (β, -0.05; 95% CI, -0.08 to -0.02; P < .001) at age 8 years (mean [SD] age at measurement, 8.6 [0.3] years) and symptoms of hyperactivity and inattention from ages 4 to 16 years, with the strongest evidence of association at age 13 years (mean [SD] age at assessment, 13.2 [0.2] years). The odds ratio at this age per SD increase in PRS was 1.25 (95% CI, 1.12-1.39) (P < .001). There was little evidence of association between the RA PRS and other measures of cognition and psychopathology. Gene-based analyses indicated that polygenic signal for RA was enriched for immune pathways (q ≤ 0.05). No equivalent associations were seen for polygenic risk associated with inflammatory bowel disease or multiple sclerosis.

Conclusions and relevance: These findings support an association between genetic risk for RA and neural phenotypes, suggesting that cognitive impairment in RA is not simply secondary to disease-related processes or treatment effects. These results may suggest that genetic susceptibility for RA might affect psychological well-being in early life and reinforce the emerging link between mental health and the immune system.

Figure 1.. Associations Between Polygenic Risk Scores (PRSs) for Rheumatoid Arthritis (RA), Inflammatory Bowel Disease (IBD), and Multiple Sclerosis (MS) and Cognitive Measures at Age 8 Years

Points represent β values, with error bars indicating 95% confidence intervals. β values are shown for PRSs generated using lists of single-nucleotide polymorphisms meeting a P value threshold of .05. The horizontal black line indicates the null value. The number of observations for total IQ was 5305; performance IQ, 5320; verbal IQ, 5328; working memory, 5210; verbal learning, 5334; processing speed, 5340; problem solving, 5282; selective attention, 5105; and attentional control, 5177. Numerical results used to generate the figure are presented in eTable 4 in the Supplement.

Figure 2.. Associations Between Polygenic Risk Scores (PRSs) for Rheumatoid Arthritis (RA), Inflammatory Bowel Disease (IBD), and Multiple Sclerosis (MS) and Hyperactive and Inattentive Symptoms Measures at Multiple Ages

Points represent odds ratios (ORs), with error bars indicating 95% confidence intervals. Odds ratios are shown for PRSs generated using lists of single-nucleotide polymorphisms meeting a P value threshold of .05. The horizontal black line indicates the null value. The number of observations for age 4 years was 5936; 7 years, 5531; 8 years, 5291; 10 years, 5551; 12 years, 5129; 13 years, 4953; and 16 years, 4089. Numerical results used to generate the figure are presented in eTable 5 in the Supplement.

Figure 3.. Associations Between Polygenic Risk Scores (PRSs) for Rheumatoid Arthritis (RA), Inflammatory Bowel Disease (IBD), and Multiple Sclerosis (MS) and Inflammatory Markers

Points represent β values for interleukin 6 (IL-6) and C-reactive protein (CRP), with error bars indicating 95% confidence intervals. β values are shown for PRSs generated using lists of single-nucleotide polymorphisms meeting a P value threshold of .05. The horizontal black line indicates the null value. The number of observations for IL-6 at age 9 years was 4055; CRP at age 9 years, 4064; and CRP at age 16 years, 2779. Numerical results used to generate the figure are presented in eTable 6 in the Supplement.