The gut-brain axis in Parkinson’s disease : epidemiological studies on causes, underlying mechanisms and novel treatments

Abstract: Parkinson’s disease (PD) is an age-related neurodegenerative disease with highly heterogeneous symptomatology. Little is known about why PD occurs. The gastrointestinal tract has been postulated as an origin of PD pathology, which then spreads along the gut- brain axis to enter the brain. This so-called “Braak hypothesis” well explains the gastrointestinal symptoms that are experienced by nearly 80% of PD patients during the prodromal stage of disease and the detectable α-synuclein (the pathogenic protein in PD) in multiple sites outside the brain. It has also been supported by subsequent evidence emerging from various settings. Hence, this thesis aimed to expand our knowledge about PD etiology and to inform therapeutic innovation from the perspective of the gut-brain axis. In Study I, we tested whether Clostridium difficile infection (CDI), an infectious diarrhea caused by a Gram-positive bacillus, is a risk factor for PD in a cohort study. By following 4.7 million Swedish adults between 1997 and 2013 for incident PD diagnosis, we found that CDI was associated with a short-term PD risk increase, which may be spurious due to reverse causation or surveillance bias, but was not associated with long-term PD risk. In Study II, we examined the relationship between microscopic colitis (MC), a chronic intestinal inflammatory disease, and PD. Using a matched cohort design, we showed that PD incidence was only associated with MC during the first five years since MC diagnosis, but not in longer follow-up. A post-hoc matched case-control analysis also found that PD prevalence was higher among MC versus MC-free individuals. Collectively, these findings imply that MC is more likely to be a comorbidity than a risk factor for PD. In Study III, we performed a Mendelian randomization study to estimate the causal effect of tumor necrosis factor (TNF) inhibition, an anti-inflammatory therapy for inflammatory bowel disease (IBD), on PD. Pharmacologic blockade of the proinflammatory signaling mediated by TNF receptor 1 (TNFR1) was genetically instrumented by four variants in the vicinity of the TNFR1-encoding gene. Results from different models consistently showed no benefits of TNF/TNFR1 antagonism for the prevention or delay of PD onset for the general population. In Study IV, genetic overlap between IBD and PD was explored via the conditional false discovery rate framework. We detected robust evidence for a genetic link between PD and each subtype of IBD, underpinned by many shared genomic loci. The genetic findings suggested the presence of both common etiology and antagonistic pleiotropy between PD and IBD risk genes, and further highlighted host immunity and/or autoimmunity in explaining the biological connection between the two seemingly unrelated diseases. In summary, findings from this thesis shed new light on the relationship between several intestinal diseases and PD, and from a public health perspective, provide novel data about the potential of anti-inflammatory treatment for intestinal disorders to prevent PD. Future work is warranted to replicate and follow-up with our findings and to translate the knowledge into clinical practice.