We recently discovered in a nationwide registry-based study that in Finland subjects diagnosed with Parkinson’s disease (PD) had a higher number of antibiotic courses in the years before diagnosis than control subjects.

So does this mean that antibiotics cause PD and that people should not use antibiotics?

The short answer is “NO”, but a bit longer answer may make the underlying connections more clear. PD is most definitely a multifactorial disease, meaning that, with the exception of rare familiar forms, no single factor is enough to cause the disease. Most likely, genetic factors (vulnerability), environmental factors and age interact and only if the aggregated pathogenetic pressure reaches a certain threshold, the pathological process of PD starts.

So, antibiotics alone definitely do not cause PD. But rather it seems that an increased cumulative use of antibiotics is one among other environmental factors that increases the risk of PD. In subjects that also have accumulation of other predisposing genetic of environmental factors, extensive antibiotic use may be a final trigger that contributes to initiation of PD.

But could the increased antibiotic use be instead a reflection of a higher number of infections before onset of PD and not itself predispose? While we cannot completely exclude this, our analysis was adjusted for the number of infections. Furthermore, it is important to note that the overall antibiotic exposure was not particularly strongly linked to PD-risk. Instead, specific groups of antibiotics were highly associated, but others not.

It turned out that the antibiotics with the strongest links to PD-risk were those that have particularly strong impact on the composition of the microbiome. These antibiotics were acting against anaerobic bacteria and had a broad spectrum. In contrast, antibiotics that do not act on anaerobic bacteria and have only a narrow spectrum, were not linked to PD. This specificity for certain antibiotics also argues against the hypothesis that simply the number of infections would be responsible for this link.

It was fascinating to see that clearly the strongest connections between antibiotics and PD were seen for exposures that happened 10-15 years before the diagnosis of PD. First, this makes it unlikely that early PD symptoms such as swallowing problems that can cause more infections would explain the higher antibiotic use. Second, it fits extremely well with the current hypothesis of PD development. PD pathology may start many years before the first motor symptoms and PD diagnosis. One point of disease initiation may be the gut.

In my opinion an intriguing explanation for our findings is that a disruption of the gut microbiome caused by increased antibiotic use may make the gut more vulnerable to the development of PD pathology. However, also other direct effects of antibiotics on nerve cells may be possible.

Many PD patients have mentioned to me that they experienced changes in their PD symptoms during or after a course of antibiotics. Both improvements and deteriorations were mentioned. For the future it will be important to further disentangle these connections to understand how antibiotics and PD are linked mechanistically.

Finally, our findings do not immediately change the principles of using antibiotics. In all cases, antibiotics should be used only in conditions where they likely benefit the patient and not for mild infectious symptoms that could also be caused by a virus. Furthermore, narrow spectrum antibiotics should be preferred, if possible. Anyway, these guidelines are essentially the same as for the prevention of bacterial resistance problems and are very important to implement. Antibiotics save millions of lives each year, but we continuously learn more and more about their impact on human health beyond the problem of antibiotic resistance.

Filip Scheperjans, MD, PhD, Adjunct professor, Department of Neurology, Helsinki University Hospital spoke at the WPC 2016 in Portland, Oregon.

Ideas and opinions expressed in this post reflect that of the author(s) solely. They do not necessarily reflect the opinions of the World Parkinson Coalition®