Causal Associations Between Imaging-derived Phenotypes and Risk of Alzheimer’s Disease and Other Neurodegenerative Disorders:  A Mendelian Randomization Study

doi:10.21203/rs.3.rs-5065314/v1

Download PDF

Research Article

Causal Associations Between Imaging-derived Phenotypes and Risk of Alzheimer’s Disease and Other Neurodegenerative Disorders: A Mendelian Randomization Study

https://doi.org/10.21203/rs.3.rs-5065314/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background

Accumulating observational studies have suggested associations between imaging-derived phenotypes (IDPs) and common neurodegenerative disorders, especially Alzheimer’s disease (AD). The goal of this study is to evaluate the causal associations between structural and functional IDPs and 4 neurodegenerative disorders, including AD, Parkinson’s disease (PD), Amyotrophic lateral sclerosis (ALS), and Multiple sclerosis (MS).

Methods

Bidirectional two-sample Mendelian randomization (MR) studies were conducted using summary statistics obtained from genome-wide association studies of 3909 IDPs from UK biobank and 4 neurodegenerative disorders.

Results

Forward MR analysis showed that volume of cerebral white matter in the left hemisphere was associated with increased risk of ALS (odds ratio [OR] = 1.15, 95% confidence interval [CI] = 1.09–1.22, P = 3.52 x 10^− 6). In reverse MR analysis, we revealed genetically determined risk of AD and MS were associated with multiple IDPs (all P < 1.28 x 10^− 5[0.05/3909], 9 IDPs in AD and 4 IDPs in MS). For example, genetically determined risk of AD was causally associated with reduced volume of gray matter in right ventral striatum (OR = 0.95, 95% CI = 0.93–0.97, P = 4.68 x 10^− 7) and lower rfMRI amplitudes in several nodes (ICA25 node 9, ICA25 node 8, and ICA100 node 11). Additionally, genetically determined risk of MS was causally associated with reduced volume in left putamen (OR = 0.97, 95% CI = 0.97–0.98, P = 4.47 x 10^− 7) and increased orientation dispersion index in right hippocampus (OR = 1.03, 95% CI = 1.01–1.04, P = 2.02 x 10^− 6).

Conclusions

Our study suggested plausible causal associations between risk of NDDs and brain IDPs. These findings might hold promise for identifying new disease mechanisms and developing novel preventative therapies for NDDs at the brain imaging levels.

Neurobiology of Disease

Alzheimer’s disease

Imaging-derived phenotypes

Mendelian randomization

Neurodegenerative disease

Neurodegenerative disorders (NDDs) are characterized by the chronic degeneration of nerve cells in central or peripheral nervous system, finally leading to motor, cognitive, emotional, and autonomic disturbances. The prevalence of NDDs is increasing in part due to the extensions of lifespan, however, there is still no cure for any of these diseases. The most common NDDs include Alzheimer’s disease (AD) and Parkinson’s disease (PD), which significantly decrease life expectancy and quality of life^{1, 2}.

All the NDDs exhibit characteristic imaging-derived phenotypes (IDPs) in magnetic resonance imaging (MRI) examinations, such as medial temporal lobe atrophy in AD and white matter lesions in MS. Indeed, a multitude of observational studies have been conducted to investigate the relationships between brain IDPs and NDDs. Previous studies have tried to identify potential IDPs that may affect the risk of NDDs. For example, specific brain atrophy patterns have been revealed to be associated with faster cognitive decline and highest risk of developing AD³. Fractional anisotropy in fornix has been demonstrated to predict the progression of mild cognitive impairment to AD in a longitudinal study⁴. Besides, specific MRI characteristics have been utilized to predict the diagnosis of multiple sclerosis (MS) in children with CNS demyelination⁵. In addition to focusing on the effects of brain IDPs on risk of NDDs, many studies have used MRI to assess the changes of brain IDPs in patients diagnosed with NDDs. For example, PD patients exhibited reduced cortical-subcortical sensorimotor connectivity compared with healthy control participants⁶. AD patients exhibited increases in cortical thickness and decreases in cortical diffusivity in early preclinical stage, but reduced cortical thickness and increased cortical diffusivity in symptomatic stage⁷. Although accumulated studies have reported potential relationships between brain IDPs and NDDs, whether brain IDPs causally affect the risk of NDDs or whether NDDs causally affect brain IDPs remain largely unknown.

Mendelian randomization (MR) is a powerful statistical method that utilizes genetic variants as instruments to infer the causal associations between exposure and outcome^{8, 9}. Considering the critical role of brain IDPs in the occurrence of NDDs and NDD-associated clinical symptoms, it is important to examine whether brain IDPs are causally associated with NDDs using MR analysis. Stephen et al. (2021) previously published genome-wide associations of 3,935 brain IDPs, thus providing a valuable source to examine the causal associations between brain IDPs and NDDs using bidirectional MR analysis. According to previous studies, the MR analyses have demonstrated causal associations between several brain IDPs and education attainment¹⁰, cognitive impairment¹¹, stroke¹², and common psychiatric disorders^{13, 14, 15}. In this study, we used two-sample bidirectional MR method to systematically investigate the causal associations between 3909 IDPs and 4 NDDs. Our study might provide new insights into the diagnosis and treatment of common NDDs.

Ethical approval

This MR study is performed based on the open-access summary statistics from previous published genome-wide association studies (GWAS) with ethical approval by corresponding ethics committee, thereby, additional ethical approval is not required.

Study design

The overall study design is illustrated in Fig. 1. MR analyses were conducted in accordance with the STROBE-MR checklist ¹⁶ and Burgess et al.’s guidelines¹⁷. In forward MR analysis, we treated each brain IDP as the exposure (a total of 3909 IDPs; sample size: n = 27663 ~ 33219)¹⁸ and the risk of each NDD (PD, AD, Amyotrophic lateral sclerosis [ALS], and MS; sample size: n = 115,803 ~ 482,730) as the outcome^{19, 20, 21, 22}. In reverse MR analysis, the risk of each NDD was set as the exposure and each brain IDP as the outcome.

Exposure data

In forward MR analysis, the GWAS data of brain IDPs were sourced from BIG40 (https://open.win.ox.ac.uk/ukbiobank/big40/)¹⁸ and the IEU OpenGWAS project. This dataset comprised the genome-wide mapping of 3935 brain IDPs derived from a sample size of 33,224 individuals of European ancestry from the UK Biobank release 2020¹⁸. To avoid meaningless statistical analyses, we excluded several IDPs from primary 3935 IDPs, such as quality control-related IDPs. We kept nearly all brain IDPs derived from different parcellation schemes, considering that parcellation methods significantly affect the imaging properties of brain IDPs. We also included several brain IDPs that are relatively less investigated according to previous literature, including IDPs based on cortical grey-white contrast and tissue intensity. The finally selected 3909 IDPs investigated in this study were listed in Additional file 1: Table S1. They were divided into 4 categories: structural MRI IDPs, diffusion MRI (dMRI)-based IDPs, functional MRI (fMRI)-based IDPs, and susceptibility-weighted imaging (SWI)-based IDPs. In reverse MR analysis, the summary statistics of 4 NDDs, including PD (GWAS ID: ieu-b-7)²⁰, AD (GWAS ID: ieu-b-2)²¹, ALS (GWAS ID: ebi-a-GCST90027163)²² and MS (GWAS ID: ieu-b-18)¹⁹, were all extracted from IEU OpenGWAS project (https://gwas.mrcieu.ac.uk/). All participants included in above original GWAS were of European ancestry. The GWAS summary statistics of 4 NDDs have been widely used in previous MR studies^{23, 24}.

Outcome data

In forward MR analysis, the GWAS summary statistics of each NDD were obtained from IEU OpenGWAS project (https://gwas.mrcieu.ac.uk/) as above described. In reverse MR analysis, the summary statistics of 3909 brain IDPs were sourced from IEU OpenGWAS project (https://gwas.mrcieu.ac.uk/).

Selection of genetic IVs

According to the standard procedures of MR studies^{8, 9, 16, 17}, each instrumental variable (IV) in this study should meet 3 assumptions as described below: (i) the IV significantly correlated with the exposure; (ii) the IV was independent of all other IVs and not associated with confounders; (iii) the IV affected the outcome only via the exposure. During the MR analysis, multiple statistical approaches such as inverse variance weighted (IVW) method and weighted median method were performed, but only IVW was selected as the main method to infer causal relationships between the exposure and outcome due to its high statistical power^{13, 23}.

To increase the statistical power of the forward MR analysis, our study initially used a relaxed statistical P-value threshold (P < 1 x 10^− 5) to screen for IVs as previously described¹⁴. Then, the linkage disequilibrium (LD) clumping was further applied within a 10 MB window using 1000 Genomes Project Phase 3 reference panel for the European populations to identify SNPs that were independently (r² < 0.001) associated with brain IDPs. In the reverse MR analysis, P < 5 x 10^− 8 and r² < 0.01 were set to screen the IVs for each NDD.

If a particular requested SNP was not present in the outcome GWAS, then highly correlated proxy SNPs (r² > 0.8) were searched as IVs. F-statistics were performed to measure the power of IVs and F-statistic < 10 indicates greater bias.

Removing confounders

SNPs significantly associated with confounding factors, such as drinking and smoking, were excluded. Specifically, PhenoScanner V2 (http://www.phenoscanner.medschl.cam.ac.uk/) and NHGRI-EBI GWAS catalog (https://www.ebi.ac.uk/gwas/docs/file-downloads/) were used to exclude the SNPs associated with confounders.

Two-sample MR analyses

Initially, an IVW regression method with a fixed-effects model was utilized as the primary causal inference. If the heterogeneity test is significant, a random-effects model was used during IVW regression. Complementary MR methods included MR Egger, weighted median, simple mode, and weighted mode, which might strengthen the reliability of the IVW estimates. In forward MR analysis, MR-RAPS (Robust adjusted profile score)²⁵ was also used to measure the robustness of the IVW estimates because of the usage of relatively weak IVs²⁵. These statistical methods were realized by the functions ‘mr_ivw’, ‘mr_egger_regression’, ‘mr_weighted_median’, ‘mr_raps’, and ‘mr_weighted_mode’ in the TwoSampleMR v0.4.26 R package. Bonferroni-corrected P < 0.05 (uncorrected P < 1.28 x 10^− 5 [0.05/3909]) was considered statistically significant. Bonferroni-corrected P < 0.05 within each category of IDPs was suggested to have potential significance.

Sensitivity analysis

First, heterogeneity test using Cochran’s Q and Rucker’s Q′ statistics²⁶ and funnel plot were utilized to assess the heterogeneity of IVs. Second, leave-one-out (LOO) analysis was conducted to assess whether the causal association was mainly driven by a single IV and IVs with P < 0.05 were regarded outliers. Third, MR-PRESSO global test (https://github.com/rondolab/MR-PRESSO/) was utilized to detect IVs having horizontal pleiotropy. Finally, an MR-Egger regression was conducted to detect the potential bias caused by directional pleiotropy. The intercept in the Egger regression indicated the existence of directional pleiotropy when the value differed from zero.

Data availability

GWAS statistics of brain IDPs were collected from BIG40 web browser (https://open.win.ox.ac.uk/ukbiobank/big40).

Code availability

All software packages we used in the study are publicly available, and the download links are included in the “Methods” section. The Code was available from the corresponding author upon reasonable request.

Forward MR: the putative causal effects of IDPs on NDDs

In the forward MR analysis, we found no brain IDPs were causally associated with the risk of AD, PD, and MS (all P > 1.28 x 10^− 5; Additional file 1: Fig.S1-3). However, several brain IDPs might exhibit suggestive associations with above diseases (P < 1.00 x 10^− 4). For example, higher cortical thickness in superior medial temporal pole of left hemisphere was potentially associated with reduced risk of AD (OR = 0.68, 95%CI = 0.56–0.82, P = 5.96 x 10^− 5; Additional file 1: Fig. S1). Additionally, greater area in left lateral orbitofrontal cortex might be associated with a lower risk of MS (OR = 0.79, 95%CI = 0.71–0.88, P = 4.54 x 10^− 5; Additional file 1: Fig. S3). Importantly, we revealed the higher volume of cerebral white matter in the left hemisphere was causally associated with higher risk of ALS (odds ratio [OR] = 1.15, 95% confidence interval [CI] = 1.09–1.22, P = 3.52 x 10^− 6; Fig. 2), which passed the sensitivity analysis.

Reverse MR: the putative causal effects of NDDs on IDPs

In the reverse MR analysis, we found genetically determined risk of AD and MS were causally associated with multiple brain IDPs (all P < 1.28 x 10^− 5, 9 IDPs in AD and 4 IDPs in MS). Particularly, in structural MRI IDPs, genetically determined risk of AD was causally associated with reduced volume of grey matter in right ventral striatum (OR = 1.06, 95% CI = 1.04–1.08, P = 1.72 x 10^− 9; Table 1, Fig. 3–4, Additional file 1: Fig. S4-5) and increased mean intensity in right amygdala (OR = 1.05, 95% CI = 1.03–1.07, P = 4.68 x 10^− 7; Table 1, Fig. 3–4, Additional file 1: Fig. S4-5) and right Accumbens area (OR = 1.05, 95% CI = 1.03–1.07, P = 3.33 x 10^− 7; Table 1, Fig. 3–4, Additional file 1: Fig. S4-5). The genetically determined risk of AD was also causally associated with lower intensity-contrast in right inferior temporal gyrus (OR = 0.95, 95% CI = 0.93–0.97, P = 4.54 x 10^− 6; Table 1, Fig. 3–4, Additional file 1: Fig. S4-5) and right parahippocampal gyrus (OR = 0.96, 95% CI = 0.94–0.98, P = 1.02 x 10^− 5; Table 1, Fig. 3–4, Additional file 1: Fig. S4-5). In functional MRI IDPs, genetically determined risk of AD was causally associated with lower rfMRI amplitudes in several nodes, including ICA25 node 9, ICA25 node 8, and ICA100 node 11 (Table 1, Fig. 3–4, Additional file 1: Fig. S4-5) and increased rfMRI connectivity (ICA100 edge 681; Table 1, Fig. 3–4, Additional file 1: Fig. S4-5). For MS, genetically determined risk of MS was causally associated with reduced volume in left putamen (OR = 0.97, 95% CI = 0.97–0.98, P = 4.47 x 10^− 7; Table 1, Fig. 5–6, Additional file 1: Fig.S6-7) and increased orientation dispersion index in right hippocampus (OR = 1.03, 95% CI = 1.01–1.04, P = 2.02 x 10^− 6; Table 1, Fig. 5–6, Additional file 1: Fig. S6-7) and higher area in parahippocampal gyrus in both DKTatlas (OR = 1.02, 95% CI = 1.01–1.03, P = 3.21 x 10^− 6; Table 1, Fig. 5–6, Additional file 1: Fig. S6-7) and Desikan-Killiany parcellation (OR = 1.02, 95% CI = 1.01–1.03, P = 3.37 x 10^− 6; Table 1, Fig. 5–6, Additional file 1: Fig. S6-7). The genetically determined risk of PD and ALS were not found to be causally associated with brain IDPs (all P > 1.28 x 10^− 5; Additional file 1: Fig. S8-9).

Table 1

The putative effects of NDDs on brain IDPs examined by MR analysis
Exposure (NDDs)	Outcome				OR (95%CI) ^a	p-value ^b	IV number	Sensitivity analysis ^c	Reverse effect ^d
Exposure (NDDs)	UKB ID			IDPs	OR (95%CI) ^a	p-value ^b	IV number	Sensitivity analysis ^c	Reverse effect ^d
		AD
AD	26578		aseg rh intensity Amygdala		1.06(1.04, 1.08)	1.72E-09	27	passed	No
AD	25753		rfMRI connectivity (ICA100 edge 681)		1.06(1.04, 1.08)	2.31E-08	27	passed	No
AD	26579		aseg rh intensity Accumbens-area		1.05(1.03, 1.07)	3.33E-07	27	passed	No
AD	25891		IDP T1 FAST ROIs R ventral striatum		0.95(0.93, 0.97)	4.68E-07	27	passed	No
AD	25754		rfMRI amplitudes (ICA25 node 9)		0.95(0.94, 0.97)	3.58E-06	27	passed	No
AD	27032		wg rh intensity-contrast inferiortemporal		0.95(0.93, 0.97)	4.54E-06	27	passed	No
AD	25754		rfMRI amplitudes (ICA25 node 8)		0.96(0.94, 0.97)	6.24E-06	27	passed	No
AD	27039		rfMRI amplitudes (ICA100 node 11)		0.96(0.94, 0.98)	8.01E-06	27	passed	No
AD	25755		wg rh intensity-contrast parahippocampal		0.96(0.94, 0.98)	1.02E-05	27	passed	No
		MS
MS	25015		IDP T1 FIRST left putamen volume		0.97(0.97, 0.98)	4.47E-07	82	passed	No
MS	25428		IDP dMRI TBSS OD Cingulum hippocampus R		1.03(1.01, 1.04)	2.02E-06	82	passed	No
MS	27156		aparc-DKTatlas lh area parahippocampal		1.02(1.01, 1.03)	3.21E-06	82	passed	No
MS	26736		aparc-Desikan lh area parahippocampal		1.02(1.01, 1.03)	3.37E-06	82	passed	No
Abbreviations: AD, Alzheimer’s disease; NDDs: MR: Mendelian randomization; Neurogenerative diseases; IDP: Imaging-derived phenotypes; OR: Odds ratio; UKB: UK biobank; ICA: Independent component analysis; rfMRI: Resting-state functional magnetic resonance imaging; TBSS: Tract-Based Spatial Statistics; OD: orientation dispersion index.
^aInverse variance weighted (IVW) method was selected to be the major method to examine causal associations in MR analysis.
^sStatistical significance was defined as Bonferroni-corrected p < 0.05 (uncorrected p < 1.31E-05 [0.05/3814]).
^cThe sensitivity analysis included heterogeneity test, LOO analysis, MR-PRESSO global test, and Pleiotropy test.
^dThe reverse MR analysis was conducted to examine the reverse effect.

Observational studies have reported significant relationships between brain IDPs and common NDDs; however, whether brain IDPs and NDDs are causally correlated remains poorly understood. In this study, we performed bidirectional two-sample MR analyses to systematically dissect the causal associations between 3909 IDPs and 4 NDDs. We demonstrated that AD was causally associated with 9 brain IDPs and MS was causally associated with 4 brain IDPs. We also identified one brain IDP was causally associated with increased risk of ALS. Therefore, our findings provided new insights into the bidirectional relationships between brain IDPs and NDDs.

In the forward MR analysis, we found most of brain IDPs were not causally associated with the risk of NDDs, which provided evidence that brain IDPs exerted weak effects on the prevalence of 4 NDDs. The major finding here is the causal association between greater volume of left cerebral white matter and increased risk of ALS. Consistently, Alexander et al. (2024) demonstrated that increased cerebral white matter volume was causally associated with higher risk of ALS²⁷. However, according to previous literature, ALS patients tend to exhibit impairments of white matter integrity rather than increased white matter volume. For example, ALS patients exhibited significant white matter degeneration in frontal lobe, internal capsule, brainstem and hippocampal regions^{28, 29, 30}. This discrepancy may be explained by the disease-specific white matter plasticity during childhood and adolescence²⁷. White matter volume has been shown to be significantly increased through childhood and adolescence before slowly declining after the fourth decade³¹. White matter volume reduced with ageing and correlated with cognitive impiarment³², although mechanisms underlying age-related white matter loss are still not known. Whether the white matter volume in young preclinical ALS patients carrying disease-related genetic mutations differed from control individuals remained unknown. Interestingly, presymptomatic changes in brain volume have been observed in carriers of ALS/FTD-causing MAPT and GRN mutations in early adulthood, showing higher total intracranial volume in carriers compared with noncarriers³³. However, it has been shown that ALS patients with hexanucleotide expansion in C9orf72 showed significant bilateral degenerations in axonal structures of white matter along the corticospinal tracts and in fibers projecting to the frontal lobes³⁴. Although these findings were divergent, they might suggest potential neurodevelopmental mechanisms in the initial stage of ALS occurrence. Therefore, the causal association between hemispheric white matter volume and ALS risk might be due to the differences in brain development during adolescence or childhood, which was shaped by genetic variations, age-dependent white matter plasticity, or systemic metabolic variables²⁷. Overall, our findings provided evidence that higher brain white matter volume was a potential upstream element conferring elevated risk of ALS, thereby supporting a broad concept that brain IDPs influenced the risk of NDDs.

We identified reduced cortical thickness in left superior temporal pole was potentially associated with increased risk of AD, which was consistent with a recent MR study demonstrating that atrophy of the temporal pole was associated with higher AD risk³⁵. In addition, MR analysis has shown that reduction in the surface area of left superior temporal gyrus was also associated with a higher risk of AD³⁶. Previous studies have showed that cortical thickness in temporal pole was significantly reduced in AD patients^{37, 38} and correlated with apathy of the patients³⁹. In addition, cortical thickness in temporal pole was also significantly associated with the severity of tau pathology measured with AV-1451-PET⁴⁰. Nevertheless, whether it was the cause or the consequence of AD was an open question. Here, we provided evidence that it was atrophy of left superior medial temporal pole that putatively causes AD, but not AD led to the atrophy of left superior medial temporal pole (P > 0.05 in reverse MR analysis). Therefore, future prospective studies were encouraged to elucidate the role of left temporal pole in the prediction and diagnosis of AD.

AD patients exhibited widespread changes in structural, diffusion, and functional IDPs compared to control individuals^{7, 41, 42, 43, 44}, nevertheless, whether genetically proxied AD causally shaped brain IDPs in patients remained poorly understood. Here, we revealed genetically determined risk of AD was causally associated with reduced volume of grey matter in right ventral striatum, which indicated that gray matter loss in right ventral striatum was a consequence of AD dementia³⁶. Ventral Striatum played a key role in learning and memory⁴⁵, as well as reward processing and motivated behavior^{46, 47}. It was reported that ventral striatum exhibited stronger positive functional connectivity with the ventral caudate and medial orbitofrontal cortex, which were implicated in reward processing and motivation⁴⁸. Selective loss of cholinergic neurons in the ventral striatum has been reported in AD patients⁴⁹. In addition, amyloid-β (Aβ) has been shown to induce dopamine release in ventral striatum and decrease dopamine release in the dorsal striatum⁵⁰. These findings might indicate that reward processing was impaired in AD. Indeed, impairment in reward processing has been revealed in AD mouse models^{51, 52}. Additionally, abnormal reward behavior was also observed in typical AD patients⁵³. The dysfunction of reward processing in AD patients might be associated with the significant disruptions of reward system in the brain, including ventral striatum⁴⁷, as shown by our MR analysis. We found genetically proxied AD was also associated with higher mean intensity in amygdala^{54, 55} and accumbens area^{56, 57}, which were both key components of reward system in brain. The atrophy of amygdala has been shown in patients with early AD⁵⁸ and amygdala has been shown to play a key role in the propagation of neurofibrillary tangle pathology in AD⁵⁹. In addition, amygdala atrophy was related to global cognitive functioning⁶⁰ and the impairment of amygdala-frontal circuit has been shown to mediate the association between depressive symptoms and cognitive function in AD⁶¹. How mean intensity in amygdala changed in AD remained unknown, thereby deserving to be further explored in future studies. The impairment of nucleus accumbens has been shown to mediate reward processing dysfunction in AD. For example, loss of glycine receptors in the nucleus accumbens has been shown to induce the impairment of reward processing at an early stage of the disease⁵². In addition, age-dependent DAergic neuron loss in the ventral tegmental area has been shown to lead to lower DA outflow in the nucleus accumbens shell, which contributed to the dysfunction of reward processing⁵¹. In AD patients, lower within-network and between-network functional connectivity in reward networks (i.e., nucleus accumbens and orbitofrontal cortex) has been observed⁶². Taken together, genetically proxied AD was associated with altered brain IDPs involved in reward processing, thereby providing new insights into the neural mechanisms underlying reward dysfunction in AD patients.

We found genetically determined risk of AD was causally associated with reduced intensity-contrast in both right inferior temporal gyrus and right parahippocampal gyrus. Both right inferior temporal gyrus and right parahippocampal gyrus were core nodes within default mode network, which was significantly impaired in AD patients⁶³. Patients with AD had significantly reduced gray matter volume⁶⁴ and functional connectivity⁶⁵ in the right inferior temporal gyrus compared to control individuals. In addition, white matter abnormalities and cortical thickness reduction in right parahippocampal gyrus have also been observed in AD patients^{66, 67}. Therefore, AD was causally associated with impairments of core components belonging to default mode network.

We found genetically determined risk of MS was causally associated with reduced left putamen volume, which has been demonstrated by a recent observational study showing that MS patients exhibited lower volume in left putamen compared to control individuals⁶⁸. MS was also associated with increased orientation dispersion index in right hippocampus, which was consistent with increased mean diffusivity in right hippocampus of MS patients⁶⁹, indicating white matter abnormality in right hippocampus was a key feature of MS patients. We revealed MS was associated with increased area in left parahippocampal gyrus, which was consistent with a recent study showing a causal relationship between MS and area of parahippocampal gyrus⁷⁰. Future studies were encouraged to validate these findings in MS patients.

Overall, this study used bidirectional MR method to investigate the causal associations between genetically determined brain IDPs and risk of 4 NDDs. The findings provided strong genetic evidence for possible causal links between brain IDPs and NDDs. This will help to develop better predictive imaging biomarkers and IDP-targeted interventions for NDDs at the brain-imaging level.

Financial Disclosures

There are no financial conflicts of interest to disclose.

Supplementary Information

The online version contains supplementary material.

Acknowledgments:

This work was supported by grants from National Natural Science Foundation of China (Grant No. 81873778, 82071415, and 82060213) and National Research Center for Translational Medicine at Shanghai, Ruijin Hospital, Shanghai Jiao Tong University School of Medicine (Grant No. NRCTM(SH)-2021-03).

Collaborators GBDN (2019) Global, regional, and national burden of neurological disorders, 1990–2016: a systematic analysis for the Global Burden of Disease Study 2016. Lancet Neurol 18:459–480
Collaborators G (2024) .S.D. Global, regional, and national burden of disorders affecting the nervous system, 1990–2021: a systematic analysis for the Global Burden of Disease Study 2021. Lancet Neurol 23:344–381
Planche V et al (2021) Clinical relevance of brain atrophy subtypes categorization in memory clinics. Alzheimers Dement 17:641–652
Mielke MM et al (2012) Fornix integrity and hippocampal volume predict memory decline and progression to Alzheimer's disease. Alzheimers Dement 8:105–113
Verhey LH et al (2011) MRI parameters for prediction of multiple sclerosis diagnosis in children with acute CNS demyelination: a prospective national cohort study. Lancet Neurol 10:1065–1073
Sharman M et al (2013) Parkinson's disease patients show reduced cortical-subcortical sensorimotor connectivity. Mov Disord 28:447–454
Montal V et al (2021) Biphasic cortical macro- and microstructural changes in autosomal dominant Alzheimer's disease. Alzheimers Dement 17:618–628
Emdin CA, Khera AV, Kathiresan S (2017) Mendelian Randomization JAMA 318:1925–1926
Birney E (2022) Mendelian Randomization. Cold Spring Harb Perspect Med 12
Seyedsalehi A et al (2023) Educational attainment, structural brain reserve and Alzheimer's disease: a Mendelian randomization analysis. Brain 146:2059–2074
Siedlinski M et al (2023) Genetic analyses identify brain structures related to cognitive impairment associated with elevated blood pressure. Eur Heart J 44:2114–2125
Yu K et al (2023) Assessment of bidirectional relationships between brain imaging-derived phenotypes and stroke: a Mendelian randomization study. BMC Med 21:271
Guo J et al (2022) Mendelian randomization analyses support causal relationships between brain imaging-derived phenotypes and risk of psychiatric disorders. Nat Neurosci 25:1519–1527
Zanoaga MD et al (2024) Brainwide Mendelian Randomization Study of Anxiety Disorders and Symptoms. Biol Psychiatry 95:810–817
Mu C, Dang X, Luo XJ (2024) Mendelian randomization analyses reveal causal relationships between brain functional networks and risk of psychiatric disorders. Nat Hum Behav 8:1417–1428
Skrivankova VW et al (2021) Strengthening the Reporting of Observational Studies in Epidemiology Using Mendelian Randomization: The STROBE-MR Statement. JAMA 326:1614–1621
Burgess S et al (2019) Guidelines for performing Mendelian randomization investigations: update for summer 2023. Wellcome Open Res 4:186
Smith SM et al (2021) An expanded set of genome-wide association studies of brain imaging phenotypes in UK Biobank. Nat Neurosci 24:737–745
International Multiple Sclerosis Genetics (2019) C. Multiple sclerosis genomic map implicates peripheral immune cells and microglia in susceptibility. Science 365
Nalls MA et al (2019) Identification of novel risk loci, causal insights, and heritable risk for Parkinson's disease: a meta-analysis of genome-wide association studies. Lancet Neurol 18:1091–1102
Kunkle BW et al (2019) Genetic meta-analysis of diagnosed Alzheimer's disease identifies new risk loci and implicates Abeta, tau, immunity and lipid processing. Nat Genet 51:414–430
van Rheenen W et al (2021) Common and rare variant association analyses in amyotrophic lateral sclerosis identify 15 risk loci with distinct genetic architectures and neuron-specific biology. Nat Genet 53:1636–1648
Zeng R et al (2023) Lack of Causal Associations of Inflammatory Bowel Disease with Parkinson's Disease and Other Neurodegenerative Disorders. Mov Disord 38:1082–1088
Li D et al (2024) Associations of environmental factors with neurodegeneration: An exposome-wide Mendelian randomization investigation. Ageing Res Rev 95:102254
Zhao QY, Wang JS, Hemani G, Bowden J, Small DS (2020) Statistical inference in two-sample summary-data Mendelian randomization using robust adjusted profile score. Ann Stat 48:1742–1769
Rucker G, Schwarzer G, Carpenter JR, Binder H, Schumacher M (2011) Treatment-effect estimates adjusted for small-study effects via a limit meta-analysis. Biostatistics 12:122–142
Thompson AG, Taschler B, Smith SM, Turner MR (2024) Premorbid brain structure influences risk of amyotrophic lateral sclerosis. J Neurol Neurosurg Psychiatry 95:360–365
Ogura A et al (2022) Fiber-specific white matter analysis reflects upper motor neuron impairment in amyotrophic lateral sclerosis. Eur J Neurol 29:432–440
Muller HP et al (2016) A large-scale multicentre cerebral diffusion tensor imaging study in amyotrophic lateral sclerosis. J Neurol Neurosurg Psychiatry 87:570–579
Ishaque A et al (2022) Distinct patterns of progressive gray and white matter degeneration in amyotrophic lateral sclerosis. Hum Brain Mapp 43:1519–1534
Lebel C et al (2012) Diffusion tensor imaging of white matter tract evolution over the lifespan. NeuroImage 60:340–352
Fletcher E et al (2013) Loss of fornix white matter volume as a predictor of cognitive impairment in cognitively normal elderly individuals. JAMA Neurol 70:1389–1395
Finger E et al (2023) Neurodevelopmental effects of genetic frontotemporal dementia in young adult mutation carriers. Brain 146:2120–2131
Wiesenfarth M et al (2023) Structural and microstructural neuroimaging signature of C9orf72-associated ALS: A multiparametric MRI study. Neuroimage Clin 39:103505
Wu BS et al (2021) Cortical structure and the risk for Alzheimer's disease: a bidirectional Mendelian randomization study. Transl Psychiatry 11:476
Wang Z et al (2024) Bidirectional two-sample Mendelian randomization analyses support causal relationships between structural and diffusion imaging-derived phenotypes and the risk of major neurodegenerative diseases. Transl Psychiatry 14:215
Racine AM, Brickhouse M, Wolk DA, Dickerson BC (2018) Alzheimer's Disease Neuroimaging, I. The personalized Alzheimer's disease cortical thickness index predicts likely pathology and clinical progression in mild cognitive impairment. Alzheimers Dement (Amst) 10:301–310
Li K et al (2022) Correlation Between Brain Structure Atrophy and Plasma Amyloid-beta and Phosphorylated Tau in Patients With Alzheimer's Disease and Amnestic Mild Cognitive Impairment Explored by Surface-Based Morphometry. Front Aging Neurosci 14:816043
Yu SY et al (2020) Clinical features and brain structural changes in magnetic resonance imaging in Alzheimer's disease patients with apathy. Aging 12:19083–19094
LaPoint MR et al (2017) The association between tau PET and retrospective cortical thinning in clinically normal elderly. NeuroImage 157:612–622
Kavcic V, Ni H, Zhu T, Zhong J, Duffy CJ (2008) White matter integrity linked to functional impairments in aging and early Alzheimer's disease. Alzheimers Dement 4:381–389
Brown C et al (2024) Medial temporal lobe gray matter microstructure in preclinical Alzheimer's disease. Alzheimers Dement 20:4147–4158
Thomas JB et al (2014) Functional connectivity in autosomal dominant and late-onset Alzheimer disease. JAMA Neurol 71:1111–1122
Velioglu HA et al (2023) A structural and resting-state functional connectivity investigation of the pulvinar in elderly individuals and Alzheimer's disease patients. Alzheimers Dement 19:2774–2789
Shu SY et al (2019) A New Neural Pathway from the Ventral Striatum to the Nucleus Basalis of Meynert with Functional Implication to Learning and Memory. Mol Neurobiol 56:7222–7233
Filimon F, Nelson JD, Sejnowski TJ, Sereno MI, Cottrell GW (2020) The ventral striatum dissociates information expectation, reward anticipation, and reward receipt. Proc. Natl. Acad. Sci 117, 15200–15208
Pool ER et al (2022) Differential Contributions of Ventral Striatum Subregions to the Motivational and Hedonic Components of the Affective Processing of Reward. J Neurosci 42:2716–2728
Li CS et al (2014) Resting state functional connectivity of the basal nucleus of Meynert in humans: in comparison to the ventral striatum and the effects of age. NeuroImage 97:321–332
Lehericy S et al (1989) Selective loss of cholinergic neurons in the ventral striatum of patients with Alzheimer disease. Proc. Natl. Acad. Sci 86, 8580–8584
Mukhin VN, Borovets IR, Sizov VV, Pavlov KI, Klimenko VM (2021) Differential Influence of Amyloid-beta on the Kinetics of Dopamine Release in the Dorsal and Ventral Striatum of Rats. Neurotox Res 39:1285–1292
Nobili A et al (2017) Dopamine neuronal loss contributes to memory and reward dysfunction in a model of Alzheimer's disease. Nat Commun 8:14727
Armijo-Weingart L et al (2024) Loss of glycine receptors in the nucleus accumbens and ethanol reward in an Alzheimer s Disease mouse model. Prog Neurobiol 237:102616
Chokesuwattanaskul A et al (2023) The architecture of abnormal reward behaviour in dementia: multimodal hedonic phenotypes and brain substrate. Brain Commun 5:fcad027
Sias AC et al (2024) Dopamine projections to the basolateral amygdala drive the encoding of identity-specific reward memories. Nat Neurosci 27:728–736
Amaya KA, Teboul E, Weiss GL, Antonoudiou P, Maguire JL (2024) Basolateral amygdala parvalbumin interneurons coordinate oscillations to drive reward behaviors. Curr. Biol. 34, 1561–1568 e1564
Kim YJ et al (2024) The ion channel TRPA1 is a modulator of the cocaine reward circuit in the nucleus accumbens. Mol Psychiatry
Muir J et al (2024) Sex-biased neural encoding of threat discrimination in nucleus accumbens afferents drives suppression of reward behavior. Nat Neurosci
Padulo C et al (2023) Atrophy of specific amygdala subfields in subjects converting to mild cognitive impairment. Alzheimers Dement (N Y) 9:e12436
Stouffer KM et al (2024) Amidst an amygdala renaissance in Alzheimer's disease. Brain 147:816–829
Poulin SP et al (2011) Amygdala atrophy is prominent in early Alzheimer's disease and relates to symptom severity. Psychiatry Res 194:7–13
Du Y et al (2024) The mediating effect of the amygdala-frontal circuit on the association between depressive symptoms and cognitive function in Alzheimer's disease. Transl Psychiatry 14:301
Wang D, Belden A, Hanser SB, Geddes MR, Loui P (2020) Resting-State Connectivity of Auditory and Reward Systems in Alzheimer's Disease and Mild Cognitive Impairment. Front Hum Neurosci 14:280
Yu M, Sporns O, Saykin AJ (2021) The human connectome in Alzheimer disease - relationship to biomarkers and genetics. Nat Rev Neurol 17:545–563
Wan K et al (2023) Brain Gray Matter Volume Mediated the Correlation Between Plasma P-Tau and Cognitive Function of Early Alzheimer's Disease in China: A Cross-Sectional Observational Study. J Alzheimers Dis 92:81–93
Tang F et al (2021) Differences Changes in Cerebellar Functional Connectivity Between Mild Cognitive Impairment and Alzheimer's Disease: A Seed-Based Approach. Front Neurol 12:645171
Oliveira PP Jr. et al (2010) Use of SVM methods with surface-based cortical and volumetric subcortical measurements to detect Alzheimer's disease. J Alzheimers Dis 19:1263–1272
Yin RH et al (2015) Multimodal Voxel-Based Meta-Analysis of White Matter Abnormalities in Alzheimer's Disease. J Alzheimers Dis 47:495–507
Eshaghi A et al (2014) Temporal and spatial evolution of grey matter atrophy in primary progressive multiple sclerosis. NeuroImage 86:257–264
Valdes Cabrera D et al (2023) High-resolution diffusion tensor imaging and T2 mapping detect regional changes within the hippocampus in multiple sclerosis. NMR Biomed, e4952
Sun D et al (2024) Causal relationship between multiple sclerosis and cortical structure: a Mendelian randomization study. J Transl Med 22:83

The authors declare no competing interests.

Supplementarymaterials.pdf
Supplementary materials

Download PDF

Version 1

posted

You are reading this latest preprint version

Causal Associations Between Imaging-derived Phenotypes and Risk of Alzheimer’s Disease and Other Neurodegenerative Disorders: A Mendelian Randomization Study

Status:

Version 1

Abstract

Figures

Background

Methods

Study design

Exposure data

Outcome data

Selection of genetic IVs

Removing confounders

Two-sample MR analyses

Sensitivity analysis

Data availability

Code availability

Results

Forward MR: the putative causal effects of IDPs on NDDs

Reverse MR: the putative causal effects of NDDs on IDPs

Discussion

Declarations

Financial Disclosures

Acknowledgments:

References

Additional Declarations

Supplementary Files

Status:

Version 1