Selected Publications
*Shared first/last authors, Working group members
Older adults (n = 745) with high fecal calprotectin levels showed a striking gut microbiota profile: pro-inflammatory bacteria increased, while protective short-chain-fatty-acid producers declined. At the same time, their blood contained higher levels of inflammation markers (IL-17C, CCL-19) and the toxic metabolic product indoxyl sulfate; overweight and heart attacks were also more frequent. Elevated calprotectin could therefore be an early warning signal of common inflammatory and metabolic disorders in later life.
A Bayesian network trained on ten years of TREND cohort data uncovered fine-grained interactions among 18 risk and prodromal markers of Parkinson’s disease—for example, from age to subclinical motor impairment or from depression to sleep and circulatory problems. Subthreshold parkinsonism, sex, and an abnormal transcranial sonography of the substantia nigra were directly linked to a later diagnosis. The model therefore provides a map of marker relationships and enables the generation of realistic synthetic datasets for more precise predictions.
The investigation of 666 older participants showed that Parkinson’s risk factors—such as physical inactivity, occupational solvent exposure, and especially constipation—were linked to lower gut-microbiome diversity and a Firmicutes-rich enterotype, whereas a Prevotella-rich enterotype was associated with less frequent constipation and less subthreshold parkinsonism. By contrast, no microbiome associations could be demonstrated for olfactory dysfunction, depression, orthostatic hypotension, or an abnormal transcranial sonography of the substantia nigra. Overall, the study supports the hypothesis that specific bacterial patterns are involved in Parkinson’s development even in the prodromal phase, but it also underscores the need for prospective (multi)-omics analyses to clarify causal relationships.
Among older adults (about n=900), the share with clinically relevant depression rose only slightly from before the pandemic to three months after its onset. However, in risk groups characterized by low resilience, loneliness, high stress levels, or a history of depression, prevalence nearly doubled. These increases emerged early and persisted for 20 months. Prevention and counseling initiatives should therefore be specifically tailored to these particularly vulnerable older individuals.
In a five-year sensor study with semi-annual assessments, early-stage Parkinson’s patients showed a steady rise in step count and stride-time variability during normal walking, while controls remained almost stable. In later disease stages or at faster walking speeds, these measures changed less meaningfully. Wearable sensors can therefore capture everyday gait characteristics as sensitive long-term markers of disease progression.
Using health-insurance records from 3.7 million beneficiaries, the authors found a markedly higher prevalence and incidence of Parkinson’s disease than previously assumed (age-standardised prevalence ≈ 511 per 100 000; incidence ≈ 84 per 100 000). Affected individuals had more comorbidities (e.g., dementia, depression) and nearly twice as many physician visits, hospital admissions, and prescriptions as age-matched controls—pointing to a previously underestimated need for healthcare resources and associated costs.