Systematic Assessment of Medical Diagnoses Preceding the First Diagnosis of Multiple Sclerosis

Objective To explore the occurrence of diseases and symptoms in the 5 years before diagnosis in patients with multiple sclerosis (MS) in a case-control study.

Methods Using ambulatory claims data, we systematically assessed differences in the occurrence of diseases and symptoms in the 5 years before first diagnosis in patients with MS (n = 10,262) compared to patients with 2 other autoimmune diseases, Crohn disease (n = 15,502) and psoriasis (n = 98,432), and individuals without these diseases (n = 73,430).

Results Forty-three ICD-10 codes were recorded more frequently for patients with MS before diagnosis compared to controls without autoimmune disease. Many of these findings were confirmed in a comparison to the other control groups. A high proportion of these ICD-10 codes represent symptoms suggestive of demyelinating events or other neurologic diagnoses. In a sensitivity analysis excluding patients with such recordings before first diagnosis, no association remained significant. Seven ICD-10 codes were associated with lower odds ratios of MS, 4 of which represent upper respiratory tract infections. Here, the relations with MS were even more pronounced in the sensitivity analysis.

Conclusions Our analyses suggest that patients with MS are frequently not diagnosed at their first demyelinating event but often years later. Symptoms and physician encounters before MS diagnosis seem to be related to already ongoing disease rather than a prodrome. The observed association of upper respiratory tract infections with lower odds ratios of MS diagnosis suggests a link between protection from infection and MS that, however, needs to be validated and further investigated.