Abstract
Background
Alzheimer's disease (AD) is marked by amyloid-β and tau accumulation, processes increasingly linked to impaired protein clearance and neuroinflammation. The choroid plexus (CP), which regulates cerebrospinal fluid (CSF) production and immune signaling, may contribute to these mechanisms. This review aimed to evaluate alterations in CP volume (CPV) in AD and their clinical significance.
Methods
PubMed, Embase, Scopus, and Web of Science were searched up to March 2025. Eligible MRI-based studies comparing CPV between AD patients and healthy controls (HCs), as well as investigations examining associations of CPV with demographic, cognitive, structural, and pathological variables, were included. Random-effects models estimated pooled effect sizes, while narrative synthesis explored associations with clinical and pathological features.
Results
Sixteen studies (2004 AD; 883 HCs) met inclusion criteria. Pooled findings demonstrated significantly larger CPV in AD relative to HCs (SMD = 1.05, 95% CI: 0.67 to 1.43; p < 0.01). Narrative review indicated consistent links between CP enlargement and worse cognition, hippocampal and cortical atrophy, ventricular expansion, and increased amyloid and tau deposition. CP changes were also associated with impaired glymphatic clearance and systemic inflammation. Notably, enlargement was observed in mild cognitive impairment, suggesting early involvement in the disease course.
Conclusions
CP enlargement may represent a neuroimaging feature of AD, reflecting the interplay between impaired clearance mechanisms and neuroinflammatory processes. Given its visibility on routine MRI, CPV may hold considerable potential as an imaging marker for disease stratification and longitudinal monitoring of AD.
Introduction
Alzheimer's disease (AD) is a progressive neurodegenerative disorder and the most common cause of dementia worldwide, accounting for approximately 70% of all dementia cases. 1 Today, around 7.2 million people aged 65 and older in the United States are living with dementia due to AD, which may rise to 13.8 million by 2060. 2 AD is characterized by extracellular amyloid-β (Aβ) plaques and intracellular tau neurofibrillary tangles. 3 Recent studies show that impaired clearance of these proteins can have an important role in AD pathogenesis.4,5 Multiple pathways normally contribute to waste elimination in the brain. These pathways include transport across the blood-brain barrier (BBB), drainage through the glymphatic system, and removal at the blood-cerebrospinal fluid (CSF) barrier.4,6,7 Therefore, dysfunction across these systems leads to the accumulation of Aβ and tau, neuroinflammation, and subsequent neuronal injury in AD. Tight junctions join adjacent choroid plexus (CP) epithelial cells to form the blood-CSF barrier, which prevents paracellular free passage of molecules from the systemic circulation into the CSF. 8 Accordingly, exchange at the blood-CSF barrier is not solely explained by passive diffusion, but also involves specific exchange mechanisms across the CP epithelial interface. 9 In AD, clearance of Aβ from the brain has been described to occur via active transport at the BBB and the blood-CSF barrier. 10 Beyond transporter-mediated exchange, the CP epithelium is described as a neuroimmune interface that serves as both a physical barrier between blood and CSF and a gateway allowing peripheral immune cell entry into the CNS. 11
For these reasons, the CP has gained attention as a critical structure in AD. The CP is located within the brain's ventricles and is a highly vascularized structure that produces the majority of CSF and forms the blood–CSF barrier.4,6,7 Beyond CSF production, the CP contributes to brain homeostasis. 12 Therefore, disruption of these functions could have major consequences for CSF protein clearance and the regulation of inflammation in AD.
Previous studies show that the CP has a direct role in Aβ elimination. 13 Because the CP facilitates Aβ transport from the CSF, dysfunction can reduce clearance and accelerate plaque accumulation. Human postmortem studies show structural changes in the CP epithelium and accumulation of Aβ deposits within the CP in AD. 14 Hence, CP impairment through reduced CSF turnover and disrupted clearance mechanisms could be an important reason for AD progression.
In recent years, neuroimaging studies have assessed CP changes in AD. Several MRI-based studies have shown increased CP volume (CPV) in AD patients compared to age-matched healthy controls (HCs).15,16 Other studies have shown that CP enlargement is evident in the earlier stages of the disease. Specifically, increased CPV has been observed in individuals with mild cognitive impairment (MCI). 17 Accordingly, CP changes are present across the clinical stages of AD and may be a sensitive marker of disease activity.
Studies have reported different results, and it is unclear whether CP changes are specific to AD or reflect more general aging-related or neuroinflammatory processes. In addition, the extent of CP changes and their relationship to cognitive decline and pathology have not been systematically studied. The role of CP enlargement in AD pathophysiology and its potential as a neuroimaging biomarker requires clarification. Given the increasing recognition of the CP as both a contributor to AD mechanisms and a potential imaging marker, this review was designed with two principal objectives: (1) to compare CPV between AD patients and HCs, and (2) to evaluate the associations of CPV with demographic, cognitive, structural, and pathological characteristics in AD patients to elucidate its potential role as a biomarker of disease mechanisms and progression.
Methods
Methodological procedures of this review followed the Cochrane Handbook of Systematic Review and Meta-analysis guidelines, 18 while reporting adhered to PRISMA guidelines. 19 The review protocol was preregistered with PROSPERO (CRD420251121547).
Search strategy
Major databases, including PubMed/MEDLINE, Embase, Scopus, and Web of Science, were systematically searched for relevant studies up to March 28, 2025. To ensure a comprehensive review, additional relevant articles were identified through manual searches in Google Scholar and by screening reference lists of included studies and relevant reviews. To retrieve relevant studies, the search strategy combined keywords and corresponding MeSH terms for “Alzheimer's disease” and “choroid plexus.” The Supplementary Material outlines the detailed search strategy, with the necessary modifications implemented for each database. Given the emerging and methodologically heterogeneous nature of CP volumetric research in AD, the search strategy was intentionally designed to maximize sensitivity. Although this approach may have increased redundancy and the screening burden, it was adopted to minimize the risk of missing potentially eligible studies.
Selection process
Study selection was performed in two sequential phases. In the initial screening, titles, abstracts, and keywords were assessed to eliminate irrelevant records; full texts were obtained when abstracts were unavailable. Studies were categorized as eligible, ineligible, or potentially eligible. In the full-text review, articles deemed eligible or potentially eligible were evaluated in detail against the inclusion criteria. Two reviewers (MYP and DD) independently conducted the study selection process. Inclusion required consensus between reviewers; discrepancies were resolved through discussion, and unresolved disagreements were adjudicated by a third reviewer (OM).
Eligibility criteria
The following criteria were applied to determine study eligibility for inclusion: (1) published as peer-reviewed articles; (2) written in English; (3) cross-sectional, case-control, or cohort designs; (4) involved adult participants (≥18 years); (5) quantified CPV using structural MRI with validated acquisition protocols; and (6) either compared CPV between AD patients and HCs or examined its associations with demographic, clinical, psychological, or imaging characteristics in AD.
Exclusion criteria encompassed reviews, case reports, case series, conference abstracts, letters to the editor, books, notes, in vivo and in vitro articles, animal studies, and empirical studies that lacked sufficient methodological transparency or failed to provide access to the full text and complete dataset, despite attempts to contact corresponding authors.
Data extraction
Data extraction was performed independently by two reviewers (MYP and DD) using a standardized form that captured key information, including publication year and country, study design, diagnostic criteria for AD, distribution of groups (AD and HCs), sample sizes, sex ratios, and age for each group. Additional variables included the MRI scanner type and field strength, the software and method used for CP segmentation, the MRI sequence, the CPV normalization technique, and the major findings. For quantitative analysis, statistical parameters such as means, standard deviations, standard errors, interquartile ranges, minimum and maximum values, and confidence intervals were extracted and subsequently standardized to means and standard deviations. In instances of incomplete data, corresponding authors were contacted; studies were excluded if the necessary information was not provided. Reviewers independently validated the extracted data for accuracy and coherence; any unresolved discrepancies were arbitrated by a third reviewer (OM).
Risk of bias assessment
Risk of bias was independently assessed by two reviewers (DD and FM) using the Newcastle-Ottawa Scale (NOS), 20 with disagreements resolved through consultation with the third reviewer (OM). The NOS evaluates three domains: selection (0–5 for cross-sectional; 0–4 for cohort and case-control studies), comparability (0–2), and outcome (0–3). Studies were classified as having very high (0–3), high (4–6), or low (7–10) risk of bias based on total scores. 20
Data analysis
To assess differences in CPV between AD patients and HCs, a random-effects meta-analysis was performed using the meta package in R version 4.4.0. 21 Given the variation in reported volume units (mm3 versus % of intracranial volume), standardized mean differences (Cohen's d) were computed to ensure data comparability across studies. A random-effects model was chosen to account for methodological heterogeneity across studies. The significance of between-study heterogeneity was evaluated using the Q statistic, while its magnitude was quantified by the I2 statistic, with thresholds of 25%, 50%, and 75% indicating low, moderate, and substantial heterogeneity, respectively. 22 Egger's 23 and Begg's tests, 24 along with visual inspection of the funnel plot, were employed to assess publication bias. Sensitivity analyses were conducted by sequentially excluding individual studies and recalculating pooled estimates to assess the robustness of the results. 25 To further evaluate the potential influence of individual studies, influence diagnostics were conducted using the metafor package in R. Thus, standardized residuals were examined under a random-effects model. 26 A narrative synthesis was employed to explore potential associations between CPV and clinical characteristics of AD patients. A p-value below 0.05 was considered indicative of statistical significance in all analyses.
Results
Search results
The initial search of databases yielded 2370 records. After removing duplicates, 1476 titles and abstracts were screened, resulting in 144 full-text articles being identified for further review. Ultimately, 16 studies were included in the narrative synthesis, of which six were eligible for meta-analysis. Figure 1 illustrates the article selection process.
PRISMA 2020 flow diagram of the study process.
Characteristics of the included studies
This review incorporated sixteen studies (Table 1) published between 2019 and 2025, encompassing a total of 2004 participants diagnosed with AD (53.1% female) with a mean (SD) age of 71.6 (9.1) years and 883 HCs (41.5% female) with a mean (SD) age of 69.3 (8.9) years. The included studies originated from diverse geographical locations: six studies were conducted using data from the Alzheimer's Disease Neuroimaging Initiative (ADNI),27–32 three in China,33–35 two in South Korea,36,37 two in the Czech Republic,38,39 and one each in the Netherlands, 40 Japan, 41 and the United States. 42 Study designs were primarily case-control (n = 13) and cross-sectional (n = 3). All studies employed structural MRI with 3D T1-weighted sequences for CP volumetry, using high-field MRI scanners (1.5 T, 3 T, or 7 T). The majority utilized automated segmentation tools, particularly FreeSurfer, though a few studies employed deep learning approaches or human-in-the-loop systems. CPV was normalized to global intracranial measures, including intracranial volume (ICV), total intracranial volume (TIV), estimated total intracranial volume (eTIV), total brain volume (TBV), total cerebral volume (TCV), or supratentorial volume. However, ventricular volume was not described as a standard covariate in most included studies.
The main features of the included studies.
*Median (IQR), ADNI: Alzheimer's Disease Neuroimaging Initiative, CP: Choroid Plexus, CPV: Choroid plexus volume, CPVF: Choroid plexus volume fraction, DTI-ALPS: Diffusion Tensor Imaging–Analysis along the Perivascular Space, eTIV: Estimated total intracranial volume, F: Female, ICV: Intracranial volume, M: Male, MCI: Mild cognitive impairment, MMSE: Mini-Mental State Examination, NIA-AA: National Institute on Aging – Alzheimer's Association, NINCDS-ADRDA: National Institute of Neurological and Communicative Disorders and Stroke and the Alzheimer's Disease and Related Disorders Association, NLR: Neutrophil-to-lymphocyte ratio, NR: Not reported, SD: Standard deviation, SCD: Subjective cognitive decline, TBV: Total brain volume, TCV: Total cerebral volume, TIV: Total intracranial volume.
Primary outcomes
Comparison of CPV between AD patients and HCs
A meta-analysis was conducted on six studies comparing CPV between AD patients (n = 639) and HCs (n = 479). The pooled SMD revealed a significant increase in CPV among AD patients compared to HCs, with an overall SMD of 1.05 (95% CI: 0.67 to 1.43; p < 0.01, I2 = 88%, p-heterogeneity < 0.01), indicating a large effect size (Figure 2). Sensitivity analysis using a leave-one-out approach confirmed the robustness of the findings, and all comparisons remained statistically significant (p < 0.01) (Figure 3). Publication bias was not evident, as indicated by Begg's test (p = 0.57) and Egger's test (p = 0.83), suggesting the absence of significant funnel plot asymmetry. Influence diagnostics indicated that no study exceeded the conventional threshold for extreme outliers (|z| > 3). A study by Jiang et al. 35 showed moderate influence but did not materially alter the pooled effect estimate (Supplemental Figure 1).
Forest plot of meta-analysis of pooled SMD of choroid plexus volume between AD patients and healthy controls.
Sensitivity analysis of meta-analysis of pooled SMD of choroid plexus volume between AD patients and healthy controls.
Association between CPV and clinical features of AD patients
Several included studies reported significant associations between CPV and a range of neuropathological and clinical markers relevant to AD, reinforcing the potential of CPV as a sensitive neuroimaging biomarker. Although several studies examined the relationships between CPV and clinical or biological markers of AD, the number of eligible datasets was insufficient to allow for a pooled quantitative synthesis. Therefore, a meta-analysis could not be performed for this outcome. Instead, the available findings were integrated through a narrative synthesis, summarizing reported associations between CPV and cognitive performance, structural neuroimaging measures, amyloid and tau burden, glymphatic dysfunction, and systemic inflammatory markers. Across studies, increased CPV was consistently linked to greater cognitive impairment, as evidenced by its correlation with lower Mini-Mental State Examination (MMSE) scores, 28 worse global cognition, 37 and poorer performance on verbal learning tasks. 30 Moreover, CPV showed significant relationships with volumetric indicators of neurodegeneration, including reduced hippocampal and cortical volumes,36,37 as well as increased lateral ventricular volume (LVV) and periventricular white matter hyperintensities (WMH).36,37 These findings suggest that CPV alterations may reflect both neurodegenerative atrophy and cerebrovascular pathology. Furthermore, CPV was associated with biological markers of AD pathology, including higher levels of amyloid and tau deposition, 41 and was also linked to impaired glymphatic clearance and increased peripheral inflammation. 28 Notably, while CPV was elevated in AD patients relative to HCs, no significant difference was found between AD dementia and non-dementia subgroups, suggesting that CPV alterations may precede overt dementia. 36 Taken together, these multidimensional associations support the hypothesis that CPV enlargement reflects neuroinflammatory or neurodegenerative processes central to AD pathophysiology and may represent a pathophysiologically informative imaging marker across the clinical continuum of the disease.
Risk of bias assessment
The risk of bias was assessed using the NOS. Of the 16 studies assessed, 12 demonstrated a low risk (scores 7–9), and 4 exhibited a moderate risk (scores 4–6), with no studies categorized as high risk (scores 0–3). NOS scores ranged from 4 to 9, with a mean score of 6.8, indicating that the overall methodological quality of the included studies was moderate to high.
Discussion
This systematic review and meta-analysis provides evidence that CPV is consistently enlarged in AD patients compared to healthy individuals. The pooled effect size demonstrated a substantial volumetric difference, underscoring CP alterations as a reproducible neuroimaging feature of AD. Importantly, enlargement was also observed in patients with mild cognitive impairment, suggesting that CP changes may precede the onset of overt dementia and reflect early disease mechanisms. Beyond volumetric comparisons, our narrative synthesis highlighted multidimensional associations of CPV with cognitive decline, structural brain atrophy, ventricular expansion, Aβ and tau deposition, glymphatic dysfunction, and systemic inflammatory markers. Together, these findings reinforce the view that the CP is not merely an anatomical bystander but an active contributor to AD pathophysiology through its roles in protein clearance and immune regulation.
Our findings align with recent literature highlighting structural changes of the CP in AD. Multiple MRI-based studies have now reported that AD patients exhibit an enlarged CP compared to HCs.17,43 For instance, a large cohort study found that patients with AD dementia had the highest CPV at baseline, followed by those with MCI, with both groups showing significantly larger CPV than cognitively normal controls. 15 CP enlargement may be detectable even in prodromal stages, as shown in a population-based MRI study of 2144 elderly individuals, which reported that the CPV was an independent predictor of MCI status, even after adjusting for brain atrophy measures. 17 This suggests that CP changes emerge at an early stage of AD, preceding the onset of dementia.
CPV is a gross anatomical measure, which should be interpreted primarily as a macrostructural readout rather than a direct measure of CSF physiology or barrier function. Consistent with this interpretive limitation, recent research noted that the mechanisms linking CPV to the AD spectrum remained unclear. 35 Postmortem investigations described choroid plexus pathology in AD characterized by epithelial atrophy, basement membrane thickening, and stromal fibrosis, with associated alterations in synthesis, secretory, and transport functions and decreased CSF turnover.14,44,45 Importantly, macrostructural enlargement and microstructural degeneration are not necessarily contradictory, since reviews of CP aging describe that the CP can enlarge due to dystrophic changes while histology simultaneously shows epithelial degeneration and stromal fibrosis. 46
Recent research has further investigated CP function in AD beyond volumetric measures. MRI studies indicate that AD is accompanied by both an expansion of CSF-containing spaces (including enlargement of the CP and ventricles) and impaired glymphatic clearance of metabolites. In a study, AD patients exhibited significantly higher total CPV and ventricular CSF volume, coupled with a markedly lower DTI-ALPS index (a diffusion MRI measure of glymphatic function), compared to all other groups. 16 In that study, CPV showed an inverse correlation with the glymphatic index (larger CP associated with lower DTI-ALPS values) and a positive correlation with total CSF volume. 16 An enlarged CP in AD has been reported to co-occur with measures suggestive of altered CSF circulation. Similarly, another analysis of ADNI data found that AD patients had larger CPV and worse white-matter integrity (higher mean diffusivity) alongside lower glymphatic efficiency, and that CP enlargement was independently associated with worse cognition both cross-sectionally and longitudinally. 47 Overall, these complementary findings from imaging and neuropathological studies reinforce that CP alterations represent a reproducible structural finding in AD cohorts. However, volumetric measures alone cannot establish functional impairment or causal contribution to disease mechanisms. Notably, the substantial between-study heterogeneity observed in our meta-analysis likely reflects methodological and biological variability across studies. Differences in normalization strategies, variations in segmentation workflows, and manual quality control procedures, may contribute to variability in reported effect sizes. Furthermore, CP enlargement in AD has been reported in association with impaired glymphatic function 28 and higher peripheral systemic inflammation, 48 suggesting that volumetric expansion may reflect inflammatory and clearance-related mechanisms. Taken together, these findings suggest that CPV may reflect multiple underlying pathophysiological processes in AD rather than representing a single, uniform biological construct.
Possible mechanisms of CP enlargement in AD
The mechanisms underlying CP enlargement in AD are likely multifactorial. One prominent hypothesis is that neuroinflammation within the CP contributes to its enlargement. 49 The CP is a neuroimmune interface that houses macrophages, dendritic cells, and other immune elements, and in AD it may become activated by circulating or central inflammatory signals. 50 In line with this, evidence from both AD and other neuroinflammatory conditions suggests that CP enlargement often accompanies immune cell infiltration. For instance, a neuroimaging study found that amyotrophic lateral sclerosis patients with active inflammation had enlarged CP tissue. 51 In AD, neuropathological studies have indeed observed an accumulation of immune and inflammatory cells in the CP, along with disruption of the blood-CSF barrier. It is therefore possible that AD-related neuroinflammation induces a pro-inflammatory state in the CP, causing barrier breakdown and swelling of the plexus. An enlarged, leaky CP may permit greater entry of peripheral immune cells or substances into the CSF, further worsening neuroinflammation in a vicious cycle.
Another mechanism is that CP enlargement may represent a compensatory response to AD pathology, particularly aimed at Aβ clearance. 36 The CP produces the majority of CSF, which in turn participates in waste clearance via the glymphatic system. 52 An emerging theory suggests that if Aβ clearance is impaired in AD, the CP might hypertrophy to increase CSF production and flush out accumulated metabolites. 53 Recent data suggest that the CP supports glymphatic clearance by increasing CSF flow, with larger CPV in humans linked to greater CSF volume and altered glymphatic function.28,48 In AD patients, CPV is positively correlated with total ventricular CSF volume, hinting that an enlarged CP could be driving or responding to increased CSF production. 48 These CP changes may represent a compensatory response to facilitate the clearance of Aβ and tau, which otherwise accumulate in the brain. At the same time, glymphatic dysfunction could itself drive CP alterations. Impaired interstitial fluid drainage may lead to the accumulation of metabolites, which can secondarily damage the CP or disrupt its function.54,55
CPV changes across neurological disorders
Although CP alterations are significant in AD, they are not unique to it. Research on Parkinson's disease has reported increased CPV. A study indicated that larger CP size was linked to lower CSF α-synuclein levels, which could suggest the role of the CP in trapping pathological proteins in Parkinson's disease. 13 Multiple sclerosis also exhibits significant CPV changes. MRI studies have revealed that multiple sclerosis patients have larger and inflamed CP, 56 and the increase in CPV is detectable, as early as the radiologically isolated syndrome stage compared to HCs. 57 Moreover, the increase in CPV has also been observed in other conditions, such as schizophrenia. 58 Taken together, these findings indicate that changes in the CP may not be unique to AD, and within the current biomarker framework for AD, CPV is not considered a disease-specific marker like amyloid or tau. 59 Instead, CP enlargement probably reflects wider issues such as neuroinflammation, changed CSF dynamics, and blood-CSF barrier problems.56,60
Limitations
This study has some limitations that warrant consideration. Substantial methodological heterogeneity across included studies, such as differences in MRI scanner strength, acquisition protocols, segmentation approaches, and inclusion of distinct ventricular regions, may have contributed to variability in effect sizes despite the use of random-effects models. The predominance of cross-sectional designs limits causal inference, as it remains unclear whether CP enlargement represents a driver, compensatory mechanism, or secondary marker of AD pathology, while longitudinal data remain scarce. The interpretation of CP volumetry is further constrained by potential partial-volume effects related to ventricular enlargement and by variability in segmentation validity across automated pipelines. Differences in normalization approaches across studies may also affect comparability of volumetric estimates. Inconsistent control of comorbidities such as vascular disease and systemic inflammation may also have biased volume estimates. Interpretation is constrained by reliance on structural MRI measures alone, as volumetric changes do not necessarily reflect functional alterations. In addition, due to the limited number of eligible studies investigating associations between CPV and clinical, cognitive, or pathological factors, we were unable to perform a meta-analysis for these outcomes and instead provided a narrative synthesis. Also, restricting the review to English publications may have introduced language bias and potentially excluded relevant studies published in other languages. Moreover, the limited number of datasets eligible for quantitative synthesis precluded reliable subgroup analyses or meta-regression based on normalization strategy, segmentation method, or disease stage. Thus, we were unable to quantify the relative contribution of these factors to heterogeneity. These limitations highlight the need for standardized imaging protocols, longitudinal and multimodal studies, and integration with functional and biomarker data to better define the role of CP alterations in AD.
Conclusion
The present review consolidates current evidence indicating that CPV is increased in AD across its clinical continuum, including prodromal stages. Furthermore, available evidence suggests that CPV is linked to multiple clinical and pathological features of AD such as poorer cognitive performance, reduced hippocampal and cortical volumes, ventricular expansion, greater amyloid-β and tau deposition, impaired glymphatic clearance and heightened systemic inflammation, indicating that CP changes may represent an integrated marker of both neurodegenerative and neuroinflammatory processes. Accordingly, CPV may be considered an informative imaging marker in AD. Future research using standardized imaging protocols, multimodal biomarkers, and longitudinal follow-up is needed to clarify the specificity, temporal dynamics, and clinical utility of CPV as a biomarker for early detection, disease monitoring, and therapeutic evaluation in AD.
Supplemental Material
sj-docx-1-alz-10.1177_13872877261447404 - Supplemental material for Choroid plexus volume in Alzheimer's disease: A systematic review and meta-analysis
Supplemental material, sj-docx-1-alz-10.1177_13872877261447404 for Choroid plexus volume in Alzheimer's disease: A systematic review and meta-analysis by Farhad Mahmoudi, Mohammad Yazdan Panah, Danial Dehghani Firouzabadi, Saeed Vaheb, Hamed Ghoshouni, Ramon Flores Gonzalez, Vahid Shaygannejad and Omid Mirmosayyeb in Journal of Alzheimer's Disease
Footnotes
Acknowledgements
During the preparation of this work the authors used Chat-GPT in order to improve the readability and language of the manuscript. After using this service, the authors reviewed and edited the content as needed and take full responsibility for the content of the published article.
Ethical considerations
Not applicable
Consent to participate
Not applicable
Author contribution(s)
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Data availability statement
All relevant data are within the paper and its Supplemental Material.
Supplemental material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
