While observational studies have repeatedly pointed toward an association between obesity and multiple sclerosis, the nature of the studies makes it impossible to state with certainty that obesity leads to multiple sclerosis. Using a new statistical approach in an exceptionally large patient group, researchers at McGill University in Canada have proven that such a link very likely exists.
Since obesity rates are still increasing, the study, “Obesity and Multiple Sclerosis: A Mendelian Randomization Study,” published in the journal PLOS Medicine, underscores the importance of interventions that might lower the risk of multiple sclerosis.
All observational studies — comparing two groups to draw conclusions about disease risk or outcomes — suffer a significant drawback. Finding an association between two factors, as with obesity and neurological disease, does not imply that obesity increases multiple sclerosis risk.
The link between the two conditions might be explained by environmental factors not accounted for in the analysis. There is also a possibility that the causative mechanism might work in the opposite direction, meaning that early multiple sclerosis disease mechanisms, years before the onset of symptoms, may actually make people obese.
To circumvent the problem with observational studies, scientists developed what is known as Mendelian randomization analyses. The approach makes use of the fact that there are genetic variants that might be linked to a condition. Genes, unlike observable features, are randomly inherited without being altered by the environment.
If a gene variant is found to influence the risk of obesity, while also affecting the risk of developing multiple sclerosis, a statistical analysis can reveal any potential causality.
Taking a real-life analogy, ice cream sales and drowning accidents are highly associated. In this case, it is obvious that ice cream does not make people drown. Adding weather (in this case, the gene) to the analysis, would, however, show there is no real link between ice cream and drowning, but both can be explained by hot weather in the summer.
This was the approach the McGill research team used. Combining data from the largest genome-wide association studies (GWAS) exploring genetic risks for elevated body mass index (BMI) and multiple sclerosis, they had an exceptionally large set of data to support the analysis.
Taking all the 72 gene variants with a statistically significant link to high BMI, identified in the earlier BMI GWAS of 322,105 individuals, researchers then explored if these genes were also associated with 14,498 cases and 24,091 controls included in a GWAS for multiple sclerosis, complementing with another 9,772 patients and 17,376 controls from another study.
Adjusting the analysis for BMI allowed researchers to observe if there was a real association between a gene variant and both conditions, or if the gene only affected obesity, which, in turn, drove multiple sclerosis risk.
The findings provided researchers with the most solid proof so far that obesity in childhood or early adulthood might trigger multiple sclerosis development. For each increase in BMI of 4.7, the risk of developing multiple sclerosis increased by 41 percent. This roughly corresponds to a transitioning from overweight to obese.
The next research step to establish a definitive link between obesity and multiple sclerosis will be to study if interventions aiming to lower obesity would have an impact on multiple sclerosis rates.