Obesity at Onset Impacts the Early Clinical Presentation of Multiple Sclerosis

Background: There is evidence suggesting early clinical expression may predict long-term multiple sclerosis (MS) outcomes; however, little is known about factors influencing onset heterogeneity. In recent years the relationship between obesity and risk of MS is clear, however the influence on the manifestation of MS is not known. Given temporality, we hypothesize obesity negatively affects the early clinical expression of MS.

Objectives: To investigate the relationship between obesity and age of onset (AOO), number of impaired functional domains (NIFDs) at onset, and early relapse activity (ERA), accounting for other MS risk factors and comorbidity burden.

Methods: The study consisted of 1,524 persons with MS (PwMS) participating in the Accelerated Cure Project. Detailed medical history, socio-demographics, and behavioral risk factors were collected. Genotypic data were available for a subset (N=1,054). The outcomes were: 1. AOO (age of 1st symptom); 2. NIFDs (based on 31 symptoms at onset which were classified into FDs by two neurologists); and 3. ERA (number of relapses in the 2yrs after onset). For each outcome, we conducted variable selection using least absolute shrinkage and selection operator (LASSO). Variables included gender, birth year, race, disease subtype (relapsing remitting (RR) or primary progressive (PP) at onset), smoking status within 5yrs of onset, diagnosis of obesity by onset, history of infectious mononucleosis, education level (as a proxy for socioeconomic status [SES]), comorbidity burden, and HLA-DRB1*15:01. The relationships between the LASSO selected variables and their respective outcomes were investigated using linear (AOO) and negative binomial (NIFDs & ERA) regressions with robust standard errors. The ERA analysis was restricted to RRMS.

Results: Obesity was associated with all three outcomes; there will also other notable relationships here summarized:

For AOO, as expected HLA-DRB1*15:01 was associated with earlier AOO, and PPMS were older at onset compared to RRMS. Lower SES and being a smoker were significantly associated with earlier AOO. The obese were 8% older at onset (p=3x10^-4). Results did not vary by disease subtype, except among PPMS where males were 12% older than females.

For NIFDs, those with RRMS, of lower SES, and an of older AOO had increased NIFDs at onset. Obesity was associated with a 15% increase (p=0.04), while smoking was suggestive of an 8% increase (p=0.07) in NIFDs.

For ERA, the obese had a 30% increase in ERA (p<0.01). Younger AOO and lower SES were also associated with greater ERA.  There was suggestive evidence for a 13% increase in smokers (p=0.06).

Conclusions: We conducted a comprehensive investigation of early clinical expression in PwMS. There were multiple significant associations; of note, obesity was associated with older AOO and increases in NIFDs and ERA. These results warrant further investigations into the impact of this proinflammatory phenotype in MS.