Migraines, obsessive-compulsive disorder, and seasonal allergies seem to be associated with Parkinson’s disease and are likely to occur many years before the typical age of diagnosis, a large study has found.
Results of the study, “The Parkinson’s phenome—traits associated with Parkinson’s disease in a broadly phenotyped cohort,” were published in the journal npj Parkinson’s Disease.
Trying to identify the factors that influence whether a person will develop Parkinson’s disease is not a new idea. However, because Parkinson’s is a rare disease, such studies often don’t have enough participants to be sure that associations are real and not the result of chance.
To try to get around this problem, researchers behind a new study analyzed data from 13,546 Parkinson’s patients and more than one million control subjects to find traits associated with Parkinson’s development.
The team primarily used data from customers of 23andMe — a web-based company that provides genetic testing and analysis — who consented to be included in the study, and from surveys sent to and filled by various Parkinson’s patient groups.
After amassing the data, the researchers set about finding the factors — including family history, other diagnoses, environmental factors, personality traits, and medication usage — that were correlated with developing Parkinson’s.
The team analyzed 832 traits individually, since not every patient had data for every trait. This resulted in a list of a few hundred traits that showed some evidence of an association with Parkinson’s. The researchers then applied statistical tests and corrected for factors such as education and income, and ended up with 122 traits that were significantly associated with the disease, including some previously known factors. For example, people who consumed a lot of caffeine were less likely to have Parkinson’s, and those with the disease were more likely to report experiencing constipation.
The researchers also identified 42 traits that had not previously been associated with Parkinson’s. For example, those who could wiggle their ears or whistle were less likely to have the disease, whereas Parkinson’s patients were more likely to be afraid of heights, be farsighted, have seasonal allergies, and be married.
This large study was able to shed light on associations that had been suggested but for which there hadn’t been large enough sample sizes to draw a definitive connection. For example, the investigators found that those with obsessive-compulsive disorder (OCD) were more likely to develop Parkinson’s, supporting previous studies that were too small to achieve significance.
Although previous studies have yielded non-significant results due to their small sample sizes, “with 10,437 (Parkinson’s) cases in this particular regression, our study was well-powered to detect a positive association between (Parkinson’s) and OCD,” the researchers stated.
In addition, they found a significant association between Parkinson’s and migraines.
“We observed that migraine was positively associated with (Parkinson’s),” they said. “Only two cohort studies have previously been conducted, but both found that midlife migraine was associated with increased risk of (Parkinson’s disease) …. Since the average age of onset is typically much earlier for migraine than for (Parkinson’s), migraine may be a novel (Parkinson’s) risk factor.”
While previous studies have shown that Parkinson’s was positively associated with allergic rhinitis, but not asthma or hayfever, the current study found that it was also associated with allergies to plants and antibiotics, but not with allergies to food or animals.
“Understanding the biological differences between these different groups of allergies may allow us to hone in on specific immunological pathways that contribute to (Parkinson’s) risk,” the researchers said. “Seasonal allergies typically develop early in life, but further research is needed to determine whether having seasonal allergies is a risk factor for (Parkinson’s). If so, it may be possible to manipulate risk using immune-modulating drugs.”
The researchers noted that their study showed correlations, and not causation — that one thing does not necessarily cause another because the two are correlated. For example, people with Parkinson’s tend to report being more introverted, but it’s highly debatable whether the factors that make a person introverted also predispose them to Parkinson’s or if people with Parkinson’s are more likely to become introverted due to the difficulties of living with a neurodegenerative disease.
Similarly, marital status doesn’t seem likely to influence the development of Parkinson’s, although the researchers noted that spouses may notice sleep disturbances, which could increase the chances of people eventually getting diagnosed. At this point, however, these ideas are largely speculation.
Still, this study is not without value: “Now that we have identified phenotypes (traits) that are correlated with (Parkinson’s) it is also possible to test whether these phenotypes are causes or consequences of the disease,” the investigators wrote, adding that even in the absence of a known causal relationship, “phenotypes that are correlated with (Parkinson’s) can be used to predict (the disease) as long as the phenotype is enriched in (Parkinson’s disease) cases prior to diagnosis.”
In other words, having the data about factors that are associated with Parkinson’s can be a starting point for further research. It could be used to make predictions about Parkinson’s patients, particularly through the use of computer models and machine learning, which can interpret these massive datasets more efficiently than people.