To show that seasonal patterns of clinical and radiological multiple sclerosis (MS) activity persist despite the use of disease-modifying therapy in Montpellier and Boston cohorts.

Seasonal variation in relapses of untreated MS patients has been demonstrated in several studies with a peak in relapse frequency in spring and summer, and a trough in winter and autumn in both hemispheres. These findings suggest that environmental factors are key modulators of disease activity.

We studied MS patients living in the Montpellier and Boston areas, registered in the OFSEP (Observatoire Français de la Sclérose en plaques) and the comprehensive Longitudinal Investigation at the Brigham and Women’s Hospital (CLIMB) databases, respectively.

We used as primary clinical outcome the specific Monthly Relapse Rate (sMRR) defined as the number of relapses divided by the number of observations for a specific month during the follow-up period.

To evaluate whether the outcome distribution was seasonal, we fitted a cosinor model and compared it to a Generalized Linear Model using a Likelihood Ratio Test.

731 patients were included from the Boston (CLIMB) cohort and 568 from the Montpellier cohort.

The cosinor model resulted in a significantly better fit of monthly relapse rates in untreated patients in Montpellier as well as in Boston, when compared to GLM. In both cohorts, the monthly relapse rate of untreated patients showed an activity peak in March and a trough in September (p value = 0.0072 and 0.0017 respectively).

In contrast, we were not able to observe any significant seasonal pattern in the treated groups, both in Montpellier and in Boston.

We were able to reproduce clinical seasonal periodicity in untreated patients in both cohorts, with the temporal distribution across the year being similar in Montpellier and Boston. Residual disease activity under treatment did not demonstrate seasonal modulation.