The interplay of air pollution with plasma neurodegenerative markers and metabolome for dementia, Parkinson's disease and all-cause mortality risks and transitions: The UK Biobank study.
Aug 2026· Ecotoxicology and Environmental Safety· Vol 323, pp.
120657
· 0 citations· 73 references
Medicine
TL;DR
Findings highlight complex relationships linking air pollution, metabolic dysregulation, and neurodegeneration, and support integrative multi-omics approaches to identify pathways relevant to prevention of neurodegenerative diseases and premature mortality.
Abstract
We investigated the joint associations of circulating neurodegeneration markers-neurofilament light (NfL) and glial fibrillary acidic protein (GFAP)-ambient air pollution, and plasma metabolomic profiles with transitions from a healthy state to dementia, Parkinson's disease (PD), and all-cause mortality in the UK Biobank. The analytic sample included 19,645 participants aged ≥ 50 years with complete proteomic, metabolomic, and environmental data. Time-to-event analyses used Cox proportional hazards and multistate Weibull models to evaluate associations and statistical interactions across health transitions. Higher NfL concentrations were associated with increased risks of transitions from healthy to PD, dementia, and death, whereas higher GFAP concentrations were specifically associated with dementia (HR = 2.65, 95% CI: 2.17-3.25). At the nominal level, particulate matter (PM2.5, PM10) showed positive statistical interactions with NfL for mortality, while nitrogen oxides (NO₂/NOx) showed negative interactions with GFAP. There was little evidence of interaction between PM2.5 and either biomarker for dementia. Metabolomic principal components reflecting lipid, amino acid, and energy pathways were associated with variation in the relationships of NfL and GFAP with dementia and mortality and interacted with PM2.5 for PD and dementia. A branched-chain amino acid-related component showed a negative interaction with GFAP for dementia, indicating weaker associations at higher levels. These findings highlight complex relationships linking air pollution, metabolic dysregulation, and neurodegeneration, and support integrative multi-omics approaches to identify pathways relevant to prevention of neurodegenerative diseases and premature mortality.
BACKGROUND
Growth differentiation factor 15 (GDF15) is a stress-responsive cytokine involved in metabolic and inflammatory pathways. We examined the associations of plasma GDF15 with incident brain disorders and explored potential mediating pathways and causality.
METHODS
UK Biobank participants were followed for a median of 14 years. Plasma GDF15 was measured at baseline. Cox proportional hazards models assessed associations with incident brain disorders, including all-cause dementia (ACD), Alzheimer's disease (AD), Parkinson's disease (PD), anxiety, depression, sleep disorders, stroke, and epilepsy. Mediation analyses evaluated biochemical and hematological pathways, and one-sample Mendelian randomization (MR) was used to assess potential causal effects.
RESULTS
Higher GDF15 levels were associated with increased risks of overall brain disorders and all examined subtypes. In continuous analyses (per 1-unit increase in log2-transformed GDF15), hazard ratios (95% CIs) were 1.54 (1.48-1.60) for overall brain disorders, 1.98 (1.80-2.17) for ACD, 1.84 (1.60-2.10) for AD, 1.37 (1.18-1.59) for PD, 1.26 (1.16-1.37) for anxiety, 1.38 (1.29-1.49) for depression, 1.40 (1.26-1.55) for sleep disorders, 1.92 (1.81-2.05) for stroke, and 1.71 (1.46-2.01) for epilepsy (all P < 0.001). Lipid- and inflammation-related markers appeared to partially mediate these associations. High-density lipoprotein cholesterol (HDL-C) accounted for an estimated 7.51% of the association with depression and 11.47% with sleep disorders, while neutrophil count showed relatively larger mediation estimates across multiple outcomes. MR analyses did not support a direct causal effect of GDF15.
CONCLUSION
Plasma GDF15 is associated with a broad range of incident brain disorders and may act partly through lipid- and inflammation-related pathways, particularly HDL-C and neutrophil count.
Xinru Guo, Songyu Wu, Zhouyang Sun et al.· Progress in Neuro-psychophar...· 0 citations
OBJECTIVE
Amyotrophic lateral sclerosis (ALS) is a fatal neurodegenerative disease characterized by the degeneration of motor neurons in the brain and spinal cord. Most patients survive 3 to 5 years after diagnosis, although some live 10 years or more. Prognostic factors are incompletely understood, but previous case-control studies suggest that exposure to persistent organic pollutants (POPs) may be associated with ALS survival, but this has not been explored within prospective analysis of POP biomarkers. In this study, we aimed to prospectively investigate the relationship between exposure to POPs-assessed in pre-disease plasma samples-and survival with ALS in the Danish Diet, Cancer, and Health study cohort.
METHODS
We identified 166 incident ALS cases using the Danish National Patient Register via International Classification of Diseases (ICD) codes. We measured concentrations of 13 polychlorinated biphenyls (PCBs), 9 organochlorine pesticides, and 3 polybrominated diphenyl ethers in plasma samples collected at inclusion in the study (baseline). To assess the relationship between POP exposures and ALS survival, we used separate Cox models for each POP group adjusting for diagnosis age, sex, smoking status, body mass index, and marital status. We selected the most relevant POP groups for ALS survival using an elastic net penalized Cox proportional hazards regression model and then ran a model using a summary environmental risk score (ERS).
RESULTS
Single-pollutant Cox models showed that exposure to non-dioxin like PCBs, hexachlorobenzene (HCB), β-hexachlorocyclohexane (β-HCH), and chlordane compounds were associated with shorter survival. Elastic net penalized Cox proportional hazards regression selected HCB and chlordane compounds as the most relevant POPs for survival. The ERS Cox model showed a significant association with shorter survival (hazard ratio [HR]: 1.4, 95% confidence interval [CI]: 1.1-1.7). A 1 standard deviation (SD) higher ERS score was associated with an HR for death of 1.4 (95% CI: 1.1-1.7). Predicted median survival for those with an ERS score 1 SD higher than the mean was 12 months shorter (26 vs 38 months).
INTERPRETATION
Our study suggests that pre-disease exposure to some POPs such as HCB and chlordane compounds could negatively influence ALS survival. ANN NEUROL 2026.
A. Davias, P. Knekt, P. Rantakokko et al.· Annals of Neurology· 0 citations
Serial biomarker assessment provided stronger prognostic information than baseline measurements alone in clinically stable patients with chronic HF, and longitudinal increases in circulating amyloid-β peptides associated with cognitive decline, structural brain changes, and mortality.
J. Traub, Anna Frey, G. Homola et al.· ESC Heart Failure· 0 citations
Background Health-related factors may influence blood-based biomarkers (BBBM) of Alzheimer's disease (AD). In this analysis, associations between modifiable factors and plasma biomarkers of neurodegeneration were investigated across the Alzheimer's disease spectrum and in cognitively healthy controls in a cerebrospinal fluid–confirmed (CSF) cohort. Methods Plasma biomarkers included the Aβ1–42/1–40 ratio, pTau181, GFAP, NFL and ApoE4. Multiple linear regression was used to test associations with lifestyle factors (physical activity and sleep), physiological factors (including renal and lipid metabolism markers), genetic factors (APOE ε4), age and sex. Percentage effect sizes and confidence intervals were calculated. Results The study included CSF-characterized individuals with AD (mild cognitive impairment due to AD and AD dementia) and cognitively healthy controls (n = 116; mean age 71.2 years). Overall, the associations were modest, with wide confidence intervals reflecting variability in the outcomes and the limited range of predictors in this relatively healthy sample. In CSF-confirmed participants, age emerged as the most consistent predictor of plasma biomarker levels, particularly NFL and pTau181. APOE ε3/ε4 genotype was additionally associated with higher pTau181 levels. Other demographic, metabolic and lifestyle-related variables showed only weak or inconsistent associations. Conclusion To implement BBBM in broader populations, a systematic evaluation of confounders is required. As aging cohorts present with mixed pathologies, strategies to address heterogeneity will be essential. The limited number of robust associations observed suggests that plasma biomarkers are influenced primarily by age and genetic background rather than by metabolic factors in this cohort. Validation in more diverse populations remains warranted.
Carolin I. Kurz, Marleen Taute, Paulina Tegethoff et al.· Aging Brain· 0 citations
Introduction Type 1 diabetes (T1D) is associated with increased risk of dementias and other neurodegenerative diseases. Plasma biomarkers of neurodegeneration and neuroinflammation hold utility in risk prediction and diagnosis of neurodegenerative disease and other neuropathology, but there remains a paucity of data in the T1D population. This study quantified plasma levels of biomarkers of neurodegeneration and neuroinflammation [phosphorylated tau-181 (pTau-181), total tau (Tau), amyloid beta 42/amyloid beta 40 (Aβ42/40), glial fibrillary acidic protein (GFAP), and neurofilament light chain (NfL)], and tested associations with T1D and related characteristics. Methods This cross-sectional analysis of a subset of participants from the Coronary Artery Calcification in Type 1 Diabetes study included adults with and without T1D. Biomarker levels were quantified using Quanterix Simoa, then log-transformed and analyzed using linear regression models adjusted for age, body mass index (BMI), race and ethnicity, smoking status, and estimated glomerular filtration rate (eGFR). A subsequent sensitivity analysis included diabetic retinopathy (DR). Results 136 adults with T1D (n=90, age 52 years, 58% female, average HbA1c 7.7%) and without T1D (n=46, age 53 years, 57% female, average HbA1c 5.4%) were included. In fully adjusted models, T1D was associated with higher pTau-181 [estimate (pg/mL) 1.221, (95% confidence interval 1.003, 1.486) p=0.046], GFAP [1.296 (1.129, 1.489), p<0.001], and NfL [1.355 (1.162, 1.579), p<0.001], relative to adults without T1D. In the sensitivity analysis, relative to adults without T1D, T1D with moderate/severe DR was associated with higher pTau-181 [1.330 (1.063, 1.663), p=0.013], GFAP [1.369 (1.653, 1.608), p<0.001], and NfL [1.495 (1.252, 1.786), p<0.001], and T1D with none/mild DR was associated with higher GFAP [1.228 (1.046, 1.442), p=0.013] and NfL [1.235 (1.037, 1.470), p=0.018]. As compared to T1D with none/mild DR, T1D with moderate/severe DR was associated with higher NfL [1.211 (1.011, 1.451), p=0.038]. Across both models, eGFR was inversely associated with pTau-181, NfL, and Tau (p<0.05 for all). Discussion T1D is associated with a greater burden of plasma biomarkers of neurodegeneration and neuroinflammation, and microvascular complications such as DR and impaired kidney function may contribute to this burden. Further study is needed to fully characterize the interplay between these biomarkers and neuropathology in T1D.
Meghan E Pauley, Christina Coughlan, Fran Dong et al.· Frontiers in Immunology· 0 citations
INTRODUCTION
Alzheimer's Disease (AD) is a neurological condition that increases in severity from a gradual start. Previous studies have found an association between long sleep (LS) and AD. To slow its progression, it is important to recognize biomarkers associated with the early stages of both AD and LS.
METHODS
This study aimed to use two-sample Mendelian Randomization (MR) to identify blood and urine biomarkers associated with AD and LS. Thirty-five biomarkers from blood and urine were included as exposures in the MR analysis, while AD and LS were considered the outcomes. The AD dataset included 39,106 patients and 46,828 controls from Europe. Data on sleep duration were obtained from a GWAS in 34,184 cases and 305,742 controls of European ancestry. This study included 35 blood and urine biomarkers across 355,891 participants, including 318,953 white British, 23,582 non-white British, 6,019 African, and 7,338 South Asian participants from the UK Biobank.
RESULTS
Cystatin C, CRP, and Non-Albumin protein showed genetic association with AD and LS. The genetic association results after Bonferroni's correction are significant for Non-Albumin Protein in AD and LS.
DISCUSSION
This study provides further insight into the genetic associations and relationships among AD, LS, and other biomarkers; it also presents several potential candidates for AD and LS diagnosis and treatment. The opposing roles of C-reactive Protein, Cystatin C, and Non-Albumin protein in AD and LS warrant a greater exploration of their differential functions at various ages and in specific brain microenvironments in future studies.
CONCLUSION
This study provides a theoretical basis for the diagnosis and treatment of AD and LS. More experiments are needed in the future to verify this prediction.
Jian-Ran Ma, Zhen Wei, Chunshu Rong et al.· Current Alzheimer Research· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
