Association Between U‐p53AZ Blood Biomarker and White Matter Microstructure in Early Alzheimer's Disease: A Cross‐Sectional Study Using AlzoSure® Predict Authors: Hamide Nasiri, Setayesh Farahani, Mehrdad Mozafar, Yasaman Ghiasi, Atousa Moghadam Fard, Nasim Naseri, Amirhossein Khodadadi, Ali Salimi Asl, Maryam Mirzapour, Hossein Hamidi, Fatemeh Jafarirazi, Navid Dinarvand Published: 2025 ABSTRACT ABSTRACTBackground and AimsAlzheimer's disease (AD) is a progressive neurodegenerative disorder characterized by early microstructural white matter (WM) damage, often preceding clinical symptoms. While cerebrospinal fluid (CSF) biomarkers provide valuable insights into AD pathology, they are invasive and may not reflect early WM changes. U‐p53AZ, a conformational variant of p53 detected via the AlzoSure® blood assay, has emerged as a potential biomarker for early AD detection. This study aimed to investigate the association between AlzoSure® prediction scores and WM microstructural integrity in cognitively normal (CN) individuals and those with mild cognitive impairment (MCI), using diffusion tensor imaging (DTI). Secondary analyses explored parallels with established CSF biomarkers (Aβ₁₋₄₂, total tau, and p‐tau181).MethodsWe analyzed data from 106 Alzheimer's Disease Neuroimaging Initiative (ADNI) participants (46 CN, 60 MCI) with available DTI and biomarker data. AlzoSure® scores were derived from plasma samples using immunoprecipitation. DTI metrics (fractional anisotropy, mean, axial, and radial diffusivity) were extracted from atlas‐defined WM regions. Linear regression models, stratified by diagnosis and adjusted for age, gender, and polygenic hazard score, were used to examine associations between biomarkers and DTI indices. False discovery rate (FDR) correction was applied for multiple comparisons.ResultsIn the MCI group, higher AlzoSure® scores were significantly associated with reduced diffusivity metrics in the left cingulum hippocampal bundle, a region implicated in early AD‐related disconnection. Parallel associations were observed with CSF t‐tau and p‐tau181 across other WM tracts in MCI. However, no associations remained significant after FDR adjustment.ConclusionThese findings underscore the potential of U‐p53AZ, as measured by AlzoSure®, to capture early WM microstructural alterations particularly in individuals with MCI, aligning with tau‐related pathology patterns. While exploratory, the results highlight U‐p53AZ as a non‐invasive candidate biomarker for early AD‐related neurodegeneration. Larger, longitudinal studies are needed to validate its diagnostic and prognostic utility in preclinical and prodromal AD stages.Keywords: AlzoSure® Predict, diffusion tensor imaging (DTI), mild cognitive impairment (MCI), U‐p53AZ, white matter microstructure Introduction Alzheimer's disease (AD) has an estimated incidence of 11.08 per 1,000 individuals in Europe and is projected to affect 131 to 152 million people worldwide by 2050 [1, 2, 3]. In the United States, AD is officially ranked as the sixth leading cause of death based on death certification data [4, 5]. AD is a progressive neurodegenerative disorder characterized by gradual cognitive and memory decline, ultimately impairing activities of daily living and social functioning [2]. The pathogenesis of AD is multifactorial, arising from the interplay between genetic susceptibility and socioeconomic and environmental influences [6, 7]. Hallmark pathological features include the accumulation of extracellular amyloid‐β (Aβ) plaques and intraneuronal hyperphosphorylated tau tangles, predominantly affecting the neocortex and medial temporal lobe. These processes are accompanied by gliosis, neuronal and synaptic loss, including damage to presynaptic terminals, dendritic spines, and axons—collectively contributing to neurodegeneration [7, 8, 9, 10, 11]. Importantly, these neuropathological changes are known to begin years before the onset of clinical symptoms [6, 12]. Early identification during this preclinical phase offers an opportunity to enhance therapeutic outcomes and delay disease progression [13, 14, 15]. Although several biomarkers have been proposed for early AD diagnosis, current options are often invasive, costly, or lack predictive accuracy, and many fail to detect pathological changes long before symptom onset [16, 17]. Thus, there is a pressing need for a non‐invasive, accessible, and reliable blood‐based biomarker capable of identifying individuals at risk in the earliest stages of AD [13, 15, 17]. The selection of unfolded p53 conformational variant (U‐p53AZ) in this study is based on its emerging role in modulating neurodegenerative processes relevant to AD pathophysiology. Recent evidence highlights U‐P53AZ's capacity to regulate neuroinflammatory cascades and oxidative stress—both implicated in AD progression [18, 19]. Specifically, U‐P53AZ modulates the p53 signaling pathway, which is involved in neuronal apoptosis and synaptic dysfunction in AD [20, 21]. Preclinical studies further suggest that U‐P53AZ can cross the blood–brain barrier, confer neuroprotection by stabilizing mitochondrial function, and reduce tau hyperphosphorylation [22, 23]. Accordingly, UP53 has been proposed as a promising blood‐based biomarker for the early detection of AD [24]. Emerging data indicate that UP53 may differentiate AD from other forms of dementia and potentially predict disease onset well before clinical symptoms emerge [25, 26, 27]. Traditional techniques for detecting UP53, such as Western blotting, immunoprecipitation, immunofluorescence, DNA extraction and purification, and electrochemical immunosensors, are limited by scalability and accessibility [28, 29, 30], prompting interest in novel diagnostic approaches. AlzoSure® is a blood‐based diagnostic assay designed to detect the AZ284 peptide derived from the U‐P53AZ using the 2D3A8 antibody in immunoprecipitated plasma samples. The 2D3A8 antibody specifically binds to UP53 [25, 26]. Diffusion tensor imaging (DTI), a magnetic resonance imaging (MRI) technique, provides sensitive markers of microstructural white matter alterations and has emerged as a promising tool for detecting early brain changes associated with AD [31]. However, to date, no studies have evaluated whether UP53 levels measured via AlzoSure® are associated with microstructural white matter changes as revealed by DTI. Given the high sensitivity of DTI for detecting early white matter abnormalities and the mechanistic relevance of U‐P53AZ to AD, this study aims to address this gap. The primary objective was to investigate the association between AlzoSure® prediction scores—which estimate an individual's risk of developing AD based on U‐P53AZ levels—and regional white matter microstructural integrity assessed via DTI in cognitively normal (CN) individuals and those with mild cognitive impairment (MCI). A secondary objective was to contextualize these findings by comparing them with associations involving established cerebrospinal fluid (CSF) biomarkers of AD pathology, including Aβ₁₋₄₂, total tau, and phosphorylated tau (p‐tau181). We hypothesized that higher AlzoSure® scores would be associated with alterations in white matter microstructure indicative of early neurodegeneration, with stronger associations observed in the MCI group relative to CN individuals. Materials and Methods Participants and Study Population This cross‐sectional study utilized data from the Alzheimer's Disease Neuroimaging Initiative (ADNI) database (http://adni.loni.usc.edu/), a public–private partnership launched in 2003 under the direction of Principal Investigator Dr. Michael W. Weiner. The primary objective of ADNI is to evaluate whether serial MRI, positron emission tomography (PET), fluid biomarkers, and comprehensive clinical and neuropsychological assessments can reliably track the progression of MCI and early AD. Participants aged 55–90 years were included if they had undergone neuroimaging, lumbar puncture, and longitudinal clinical evaluation. Key exclusion criteria included a Hachinski Ischemic Score > 4, recent or unapproved medication changes, a Geriatric Depression Scale score ≥ 6, and fewer than 6 years of formal education or equivalent occupational history. Full inclusion/exclusion criteria are reported elsewhere [32]. MCI classification was based on one or more of the following: (i) a Mini‐Mental State Examination (MMSE) score between 24 and 30, (ii) a Clinical Dementia Rating (CDR) score of 0.5 with a memory box score ≥ 0.5, or (iii) a Wechsler Logical Memory II score below education‐adjusted cutoffs. All participants met the diagnostic criteria defined by the National Institute of Neurological and Communicative Disorders and Stroke and the Alzheimer's Disease and Related Disorders Association. For the current analysis, only CN individuals and those with MCI at baseline were included, provided both DTI indices and fluid biomarker data were available. A total of 106 participants met inclusion criteria, including 46 CN and 60 MCI individuals. Outliers, defined as values exceeding ±2 standard deviations from the group mean, were replaced with the group mean. AlzoSure® Predict Plasma Biomarker Assessment AlzoSure® Predict is a validated in vitro diagnostic assay designed to quantify the U‐p53AZ in human plasma. The assay employs liquid chromatography–tandem mass spectrometry (LC‐MS/MS) following immunoprecipitation with the 2D3A8 monoclonal antibody. It is intended for use in individuals aged ≥ 50 years to identify those at risk of progression to AD. A total of 593 blinded plasma samples from the ADNI were analyzed by Diadem SpA across eight sessions conducted between June and July 2022. Each session included ADNI test samples and three quality control (QC) samples sourced from the Australian Imaging, Biomarkers and Lifestyle (AIBL) biobank. Nine samples were excluded due to repeated assay failure. Quantification was based on detection of the AZ284® peptide using a TSQ Vantage LC‐MS/MS system, with a standard curve constructed from spiked calibrants (0.08–5 fmol/10 µL). Only results meeting stringent predefined quality control criteria were retained. These criteria included: coefficient of determination (R²) > 0.99 for calibration curves, > 90% recovery of synthetic U‐p53AZ, coefficient of variation (CV) < 10% for internal standard signals, and matrix effect deviation < 15%. Repeatability and concordance of QC sample measurements met U.S. FDA bioanalytical method validation guidelines, confirming the high precision and reliability of the assay across all testing sessions. CSF Biomarkers A total of 416 CS baseline samples from the ADNI cohort were analyzed for Aβ1–42, total tau (t‐tau), and phosphorylated tau (p‐tau181p) concentrations using the multiplex xMAP Luminex platform (Luminex Corp, Austin, TX). Biomarker quantification employed reagents from the INNO‐BIA AlzBio3 immunoassay kit (Innogenetics, Ghent, Belgium), designed specifically for research purposes. This assay utilizes capture monoclonal antibodies and detector antibodies specific to each biomarker, conjugated to distinct color‐coded beads. Calibration curves were established using buffered standards with known biomarker concentrations to ensure assay accuracy. Prior to analysis of ADNI and autopsy‐derived CSF samples, an interlaboratory validation study confirmed assay reliability, demonstrating coefficients of variation below 10% for CSF pool samples and under 7% for quality control samples. Additional validation through test‐retest analysis of 29 randomly selected samples showed strong reproducibility, with correlation coefficients (r²) of 0.98, 0.90, and 0.85 for t‐tau, Aβ1–42, and p‐tau181p, respectively [33, 34]. Further methodological details are available at http://www.adni-info.org/. Polygenic Hazard Score (PHS) In the initial phase of the International Genetics of Alzheimer's Project, genetic data from 17,008 patients and 37,154 controls were analyzed to identify common variants associated with AD at a significance threshold of p < 10⁻⁵. A Cox proportional hazards model was employed on the phase 1 data from the AD Genetic Consortia to evaluate these associations. To minimize population stratification effects, adjustments were made for sex, APOE genotype, and the top five principal genetic components. This approach identified 31 single nucleotide polymorphisms (SNPs), including the APOE ε2 and ε4 alleles, which were used to derive a polygenic hazard score (PHS). The PHS was subsequently integrated with age‐specific incidence rates from US population data to estimate the instantaneous risk of developing AD. The Cox model defined the PHS as the weighted sum of the genotypes across the 31 SNPs, where each SNP's contribution was scaled by its estimated effect size [35]. DTI Acquisition and Processing DTI data and region‐of‐interest (ROI) analyses were sourced from the ADNI. Preprocessing included skull stripping using FSL's Brain Extraction Tool (BET) [36], followed by linear registration of T1‐weighted anatomical scans to the Colin27 template using FSL's FLIRT tool [37]. The Colin27 brain template was resampled to match diffusion‐weighted imaging (DWI) resolution, and diffusion tensors were modeled for each voxel [38]. From the tensor eigenvalues (λ₁, λ₂, λ₃), scalar maps were derived, including fractional anisotropy (FA), mean diffusivity (MD), axial diffusivity (AxD), and radial diffusivity (RD). Lower FA and higher MD, AxD, and RD are indicative of white matter degeneration and demyelination. FA images were registered to the Johns Hopkins University (JHU) DTI atlas using an elastic, information‐based algorithm. Deformation fields were applied using nearest‐neighbor interpolation to preserve anatomical labels. This ensured alignment between atlas ROIs and participant DTI maps in a common stereotaxic space. Mean FA and MD values were extracted within each ROI, and FA values were additionally computed using tensor‐based and tract‐based spatial statistics (TBSS), following the ENIGMA‐DTI protocol [39]. All FA maps were registered to the ENIGMA‐DTI template in ICBM space, and skeletonized FA values were extracted from major white matter tracts. ROI‐level and skeleton‐based metrics were derived using standardized ENIGMA pipelines [http://enigma.loni.ucla.edu]. DTI measures reflect microstructural integrity of white matter. FA quantifies the directionality of water diffusion along axons, with higher values suggesting intact myelination and lower values indicating white matter injury [40, 41]. MD, AxD, and RD assess the magnitude and direction of water diffusion, with elevated values typically reflecting tissue damage, axonal loss, or demyelination [42, 43]. To ensure cross‐site reliability, rigorous quality control procedures were applied, including visual inspection for artifacts, motion correction, and exclusion of data failing predefined standards. Scanner‐specific calibration and harmonization protocols were implemented to minimize inter‐scanner variability. Statistical Analysis All statistical analyses were performed using Python, employing the following libraries: pandas (v2.1.4), statsmodels (v0.14.0), pingouin (v0.5.4), SciPy (v1.12.0), and numpy (v1.26.4). The Shapiro–Wilk test was used to assess the normality of continuous variable distributions. Continuous variables were summarized as mean and standard deviation (SD), while categorical variables were presented as frequencies and percentages. Between‐group comparisons were conducted using independent samples t‐tests for normally distributed continuous variables and the Mann–Whitney U test for non‐normally distributed variables. The chi‐square test was used to examine differences in categorical variables across diagnostic groups. To assess the associations between DTI‐derived metrics and either CSF biomarkers or AlzoSure® predictive scores, general linear regression models were fitted separately within each diagnostic group. All models were adjusted for age, gender, and PHS. Statistical 2‐tailed significance was defined as p < 0.05. To account for multiple comparisons, the Benjamini–Hochberg procedure was applied, and all reported p‐values for correlation and regression analyses are false discovery rate (FDR)–adjusted. Results Baseline Characteristics of the Study Population A total of 106 participants were included in the analysis, comprising 46 CN individuals and 60 with MCI. Baseline demographic, genetic, CSF biomarker, and cognitive characteristics are summarized in Table 1. There were no significant differences in age (p = 0.838) or years of education (p = 0.425) between the groups. Gender distribution differed significantly (p = 0.049). Moreover, APOE ε4 allele status was not observed significant difference between groups (p = 0.066). CSF Aβ₁₋₄₂ levels were lower in the MCI group than in CN, but the difference did not reach significance (p = 0.074). Similarly, CSF t‐tau and p‐tau181 concentrations did not differ significantly between groups (p = 0.205 and p = 0.235, respectively). Cognitive scores revealed substantial group differences. The MCI group had higher CDR‐SB compared to CN, and lower MMSE scores (p < 0.001). U‐p53AZ did not differ between groups (p = 0.337). Notably, PHS were significantly higher in the MCI group relative to CN (p = 0.022), indicating greater genetic risk in the MCI population. Table 1Baseline Characteristics of Study Population.CN (n = 46)MCI (n = 60) p valueAge, year72.97 ± 4.772.46 ± 6.50.838a GenderFemale25 (54.35%)20 (33.3%)0.049b Male21 (45.65%)40 (66.7%)Education, year16.7 ± 2.5416.32 ± 2.580.425a APOE ɛ4 allele031 (67.4%)29 (48.3%)0.066b 114 (30. 4%)24 (40%)21 (2.2%)7 (11.7%)CSF BiomarkerCSF Aβ1‐42, pg/mL1188.76 ± 422.321042.85 ± 438.20.074a CSF t‐tau, pg/mL239.09 ± 74.72260 ± 87.040.205c CSF p‐tau181, pg/mL21.61 ± 7.3824.93 ± 10.540.235a Cognitive assessmentsCDR‐SB0.09 ± 0.191.1 ± 0.51< 0.001a MMSE29 ± 1.2227.86 ± 1.68< 0.001a U‐p53AZ0.75 ± 0.180.75 ± 0.150.337a PHS0.09 ± 0.650.44 ± 0.790.022a Abbreviations: APOE, apolipoprotein E; Aβ₁₋₄₂, amyloid beta 1–42; CDR‐SB, Clinical Dementia Rating–Sum of Boxes; CN, cognitively normal; CSF, cerebrospinal fluid; MCI, mild cognitive impairment; MMSE, Mini‐Mental State Examination; PHS, polygenic hazard score; t‐tau, total tau; p‐tau181, phosphorylated tau at threonine 181. a Mann–Whitney U test. b Chi‐square test. c Independent t‐test. Baseline Characteristics of Study Population. Abbreviations: APOE, apolipoprotein E; Aβ₁₋₄₂, amyloid beta 1–42; CDR‐SB, Clinical Dementia Rating–Sum of Boxes; CN, cognitively normal; CSF, cerebrospinal fluid; MCI, mild cognitive impairment; MMSE, Mini‐Mental State Examination; PHS, polygenic hazard score; t‐tau, total tau; p‐tau181, phosphorylated tau at threonine 181. Mann–Whitney U test. Chi‐square test. Independent t‐test. Associations Between CSF Aβ₁₋₄₂ and DTI Metrics We investigated the relationship between CSF Aβ₁₋₄₂ levels and regional DTI measures using linear regression models, adjusted for age, gender, and PHS. Table 2 presents the significant associations prior to multiple comparison correction. In the CN group, higher Aβ₁₋₄₂ concentrations were significantly associated with increased fractional anisotropy (FA) in the left anterior corona radiata (FA_ACR_L; β = 0.92, p = 0.019). However, this association did not survive FDR correction (adjusted p = 0.992). In the MCI group, lower Aβ₁₋₄₂ levels were associated with higher mean diffusivity (MD) (MD_CGC_R; β = –0.07, p = 0.039), and greater radial diffusivity (RD) in the right cingulum (RD_CGC_R; β = –0.11, p = 0.019) and left superior longitudinal fasciculus (RD_SLF_L; β = –0.28, p = 0.052). None of these associations remained significant after FDR adjustment (Table 2). Table 2Significant Results of Linear Regression Analysis Between CSF Aβ₁₋₄₂ and DTI metrics.GroupDTI MetricsAdjusted R2 Standardized βT valueT value p valueAdjusted p valueCNFA_ACR_L0.160.923.002.450.0190.992MCIMD_CGC_R0.03‐0.071.452.120.0390.818RD_CGC_R0.05‐0.111.632.420.0190.818RD_SLF_L0.11‐0.282.611.990.0520.818Abbreviations: ACR, anterior corona radiata; CGC, cingulum; CN, cognitively normal; DTI, diffusion tensor imaging; FA, fractional anisotropy; L, left; MCI, mild cognitive impairment; MD, mean diffusivity; R, right; RD, radial diffusivity; SLF, superior longitudinal fasciculus. Significant Results of Linear Regression Analysis Between CSF Aβ₁₋₄₂ and DTI metrics. Abbreviations: ACR, anterior corona radiata; CGC, cingulum; CN, cognitively normal; DTI, diffusion tensor imaging; FA, fractional anisotropy; L, left; MCI, mild cognitive impairment; MD, mean diffusivity; R, right; RD, radial diffusivity; SLF, superior longitudinal fasciculus. Associations Between CSF t‐tau and DTI Metrics In the CN group, higher t‐tau levels were associated with lower FA in the right superior cerebellar peduncle (FA_SCP_R; β = –0.28, p = 0.038), and in the left tapetum (FA_TAP_L; β = 0.03, p = 0.030). In addition, t‐tau levels positively correlated with MD in the left superior fronto‐occipital fasciculus (MD_SFO_L; β = 1.10, p = 0.026), and RD in the left retrolenticular internal capsule (RD_RLIC_L; β = −0.36, p = 0.038) and left SFO (RD_SFO_L; β = 1.11, p = 0.022). In the MCI group, higher t‐tau levels were significantly associated with increased axial diffusivity (AD) in the left inferior fronto‐occipital fasciculus (AD_IFO_L; β = 0.49, p = 0.052), right medial lemniscus (AD_ML_R; β = 0.94, p = 0.046), and left uncinate fasciculus (AD_UNC_L; β = 0.42, p = 0.016). Additionally, reduced FA was observed in the left retrolenticular internal capsule (FA_RLIC_L; β = −0.31, p = 0.059), left superior corona radiata (FA_SCR_L; β = −0.23, p = 0.011), and bilateral uncinate fasciculus (FA_UNC_L: β = −0.32, p = 0.026; FA_UNC_R: β = 0.27, p = 0.047). Elevated MD and RD values were also associated with t‐tau in the left entorhinal cortex (MD_EC_L; β = 0.70, p = 0.052), left IFO (MD_IFO_L; β = 0.48, p = 0.039; RD_IFO_L: β = 0.38, p = 0.043), and left uncinate fasciculus (MD_UNC_L: β = 0.45, p = 0.005; RD_UNC_L: β = 0.45, p = 0.003) (Table 3). However, none of these associations survived FDR correction. Table 3Significant Results of Linear Regression Analysis Between CSF t‐tau and DTI metrics.GroupDTI MetricsAdjusted R2 Standardized βT valueT value p valueAdjusted p valueCNFA_SCP_R0.32−0.285.872.160.0380.894FA_TAP_L0.050.031.492.270.0300.894MD_SFO_L0.061.101.64−2.320.0260.894RD_RLIC_L0.16−0.363.03−2.150.0380.894RD_SFO_L0.071.111.71−2.400.0220.894MCIAD_IFO_L0.080.492.002.000.0520.774AD_ML_R0.020.941.192.050.0460.774AD_UNC_L0.060.421.712.510.0160.774FA_RLIC_L0.00−0.311.02−1.940.0590.774FA_SCR_L0.15−0.233.02−2.650.0110.774FA_UNC_L0.03−0.321.37−2.310.0260.774FA_UNC_R0.130.272.73−2.050.0470.774MD_EC_L0.030.701.302.000.0520.774MD_IFO_L0.060.481.782.130.0390.774MD_UNC_L0.100.452.252.980.0050.551RD_IFO_L0.030.381.362.080.0430.774RD_UNC_L0.110.452.463.120.0030.551Abbreviations: AD, axial diffusivity; CN, cognitively normal; DTI, diffusion tensor imaging; EC, entorhinal cortex; FA, fractional anisotropy; IFO, inferior fronto‐occipital fasciculus; L, left; MCI, mild cognitive impairment; ML, medial lemniscus; MD, mean diffusivity; R, right; RD, radial diffusivity; RLIC, retrolenticular internal capsule; SCP, superior cerebellar peduncle; SCR, superior corona radiata; SFO, superior fronto‐occipital fasciculus; TAP, tapetum; UNC, uncinate fasciculus. Significant Results of Linear Regression Analysis Between CSF t‐tau and DTI metrics. Abbreviations: AD, axial diffusivity; CN, cognitively normal; DTI, diffusion tensor imaging; EC, entorhinal cortex; FA, fractional anisotropy; IFO, inferior fronto‐occipital fasciculus; L, left; MCI, mild cognitive impairment; ML, medial lemniscus; MD, mean diffusivity; R, right; RD, radial diffusivity; RLIC, retrolenticular internal capsule; SCP, superior cerebellar peduncle; SCR, superior corona radiata; SFO, superior fronto‐occipital fasciculus; TAP, tapetum; UNC, uncinate fasciculus. Associations Between CSF p‐tau181 and DTI Metrics In the CN group, elevated p‐tau181 levels were significantly associated with decreased FA in multiple tracts, including the left genu of the corpus callosum (FA_GCC_L; standardized β = −0.65, p = 0.045), right medial lemniscus (FA_ML_R; β = 0.69, p = 0.041), right superior cerebellar peduncle (FA_SCP_R; β = −0.10, p = 0.046), and left tapetum (FA_TAP_L; β = 0.01, p = 0.040). Additionally, higher p‐tau181 levels were associated with increased MD and RD in the left superior fronto‐occipital fasciculus (MD_SFO_L: β = 1.02, p = 0.028; RD_SFO_L: β = 1.03, p = 0.024). In the MCI group, widespread associations between p‐tau181 and microstructural white matter changes were observed. Lower FA was significantly associated with increased p‐tau181 in several regions: bilateral body of the corpus callosum (FA_BCC_L: β = –0.53, p = 0.049; FA_BCC_R: β = –0.65, p = 0.032), left fornix (FA_FX_L: β = –0.12, p = 0.047), left medial lemniscus (FA_ML_L: β = 0.10, p = 0.034), left retrolenticular internal capsule (FA_RLIC_L: β = –0.43, p = 0.019), left superior corona radiata (FA_SCR_L: β = −0.24, p = 0.006), and bilateral body of the corpus callosum (FA_SUMBCC: β = –0.61, p = 0.040). Additionally, reduced FA in the right tapetum (FA_TAP_R: β = –0.49, p = 0.027) was associated with higher p‐tau181. Increased diffusivity measures were also associated with higher p‐tau181 in the MCI group. Elevated MD in the right posterior thalamic radiation (MD_PTR_R; β = 0.44, p = 0.036) and left uncinate fasciculus (MD_UNC_L; β = 0.44, p = 0.027), and increased RD in the same regions (RD_PTR_R: β = 0.47, p = 0.028; RD_UNC_L: β = 0.45, p = 0.020), were observed. Despite the broad pattern of significance, none of the associations in the groups survived FDR correction (Table 4). Table 4Significant Results of Linear Regression Analysis Between CSF p‐tau181 and DTI metrics.GroupDTI MetricsAdjusted R2 Standardized βT valueT value p valueAdjusted p valueCNFA_GCC_L0.07−0.651.842.070.0450.878FA_ML_R0.020.691.262.120.0410.878FA_SCP_R0.32−0.105.722.060.0460.878FA_TAP_L0.030.011.342.140.0400.878MD_SFO_L0.061.021.61−2.300.0280.878RD_SFO_L0.061.031.66−2.360.0240.878MCIFA_BCC_L0.14−0.532.76−2.030.0490.801FA_BCC_R0.13−0.652.58−2.220.0320.801FA_FX_L0.06−0.121.742.040.0470.801FA_ML_L0.040.101.512.190.0340.801FA_RLIC_L0.05−0.431.58−2.450.0190.801FA_SCR_L0.17−0.243.33−2.880.0060.801FA_SUMBCC0.15−0.613.00−2.120.0400.801FA_TAP_R0.09−0.492.18−2.290.0270.801MD_PTR_R0.090.442.082.160.0360.801MD_UNC_L0.030.441.372.290.0270.801RD_PTR_R0.060.471.742.280.0280.801RD_UNC_L0.040.451.492.420.0200.801Abbreviations: BCC, body of the corpus callosum; CN, cognitively normal; DTI, diffusion tensor imaging; FA, fractional anisotropy; FX, fornix; GCC, genu of the corpus callosum; MCI, mild cognitive impairment; MD, mean diffusivity; ML, medial lemniscus; PTR, posterior thalamic radiation; RD, radial diffusivity; RLIC, retrolenticular internal capsule; SCP, superior cerebellar peduncle; SCR, superior corona radiata; SFO, superior fronto‐occipital fasciculus; SUMBCC, summary of body of the corpus callosum; TAP, tapetum; UNC, uncinate fasciculus; L, left; R, right. Significant Results of Linear Regression Analysis Between CSF p‐tau181 and DTI metrics. Abbreviations: BCC, body of the corpus callosum; CN, cognitively normal; DTI, diffusion tensor imaging; FA, fractional anisotropy; FX, fornix; GCC, genu of the corpus callosum; MCI, mild cognitive impairment; MD, mean diffusivity; ML, medial lemniscus; PTR, posterior thalamic radiation; RD, radial diffusivity; RLIC, retrolenticular internal capsule; SCP, superior cerebellar peduncle; SCR, superior corona radiata; SFO, superior fronto‐occipital fasciculus; SUMBCC, summary of body of the corpus callosum; TAP, tapetum; UNC, uncinate fasciculus; L, left; R, right. Associations Between AlzoSure® Prediction Scores and DTI Metrics In the CN group, higher U‐p53AZ scores were significantly associated with lower AD in the left and right superior cerebellar peduncles (AD_SCP_L: standardized β = −0.51, p = 0.019; AD_SCP_R: β = −0.51, p = 0.030). A similar inverse relationship was observed for MD in the right superior cerebellar peduncle (MD_SCP_R: β = –0.50, t = −2.41, p = 0.021). In the MCI group, U‐p53AZ were inversely associated with diffusivity metrics in the left cingulum hippocampal bundle. Specifically, higher predictive scores were associated with reduced AD (AD_CGH_L: β = −0.15, t = –2.60, p = 0.012), MD (MD_CGH_L: β = –0.28, t = −2.96, p = 0.005), and radial diffusivity (RD_CGH_L: β = −0.34, t = −3.02, p = 0.004). Although these associations were statistically significant at the nominal level, none survived FDR correction (Table 5). Table 5Significant Results of Linear Regression Analysis Between U‐p53AZ and DTI metrics.GroupDTI MetricsAdjusted R2 Standardized βT valueT value p valueAdjusted p valueCNAD_SCP_L0.08−0.511.82−2.450.0190.997AD_SCP_R0.13−0.512.47−2.260.0300.997MD_SCP_R0.17−0.502.95−2.410.0210.997MCIAD_CGH_L0.16−0.153.46−2.600.0120.941MD_CGH_L0.19−0.283.84−2.960.0050.557RD_CGH_L0.18−0.343.76−3.020.0040.557Abbreviations: AD, axial diffusivity; CGH, cingulum hippocampal bundle; CN, cognitively normal; DTI, diffusion tensor imaging; L, left; MCI, mild cognitive impairment; MD, mean diffusivity; RD, radial diffusivity; SCP, superior cerebellar peduncle; R, right. Significant Results of Linear Regression Analysis Between U‐p53AZ and DTI metrics. Abbreviations: AD, axial diffusivity; CGH, cingulum hippocampal bundle; CN, cognitively normal; DTI, diffusion tensor imaging; L, left; MCI, mild cognitive impairment; MD, mean diffusivity; RD, radial diffusivity; SCP, superior cerebellar peduncle; R, right. Discussion We aimed to investigate whether the U‐p53AZ test could contribute to the prediction of AD‐related protein aggregates, genetic susceptibility, and brain microstructural alterations, as measured by DTI, in CN and MCI individuals. This study explored the association between U‐p53AZ, a novel and minimally invasive blood‐based diagnostic marker, and regional white matter microstructural integrity. Our main finding was that higher U‐p53AZ levels showed nominal correlations with reduced diffusivity in specific white matter tracts—notably, the superior cerebellar peduncles in CN and the cingulum hippocampal bundle in MCI—although these associations did not remain significant after correction for multiple comparisons. Similarly, we observed widespread nominal associations between CSF tau‐related biomarkers and white matter alterations, particularly in the MCI group, which also did not survive FDR correction. Interestingly, U‐p53AZ exhibited trends toward associations with reduced AxD, MD, and RD, suggesting a possible link to early microstructural changes that may not be captured by other biomarkers. Among CN participants, these associations were weaker and less consistent, indicating that U‐p53AZ may be more sensitive to early pathological processes occurring during the prodromal MCI stage rather than in preclinical AD. These findings align with prior research demonstrating early microstructural white matter changes in individuals at risk for AD. Previous studies have reported reduced fractional anisotropy (FA) and increased diffusivity in tracts such as the cingulum, uncinate fasciculus, and corpus callosum in MCI and preclinical AD populations [44, 45, 46]. In AD, white matter integrity is compromised, and DTI‐derived FA is frequently used to quantify associated microstructural damage [47, 48, 49]. Alterations in DTI measures have been linked to hippocampal volume, tau levels, and cognitive function [48, 50, 51, 52]. Notably, the cingulum hippocampal bundle—where we observed an association between U‐p53AZ and DTI metrics in MCI—has been implicated in early tau‐related neurodegeneration and episodic memory decline [53, 54]. The finding in the superior cerebellar peduncles among CN individuals, though less commonly reported, may reflect early compensatory or cognitive reserve mechanisms, as suggested in previous cerebellar studies [55]. Additionally, our observation that tau‐related biomarkers were more strongly associated with white matter changes than Aβ₁₋₄₂ supports the hypothesis that tau pathology is more directly linked to axonal degeneration [56, 57]. Early and accurate identification of AD in its preclinical and prodromal stages can significantly enhance treatment efficacy and clinical outcomes [13, 14, 15]. Mounting evidence suggests that AD neuropathology begins years before clinical symptoms emerge [6, 12]. While several established diagnostic tools exist—including amyloid‐PET, tau‐PET, and CSF/plasma biomarkers such as Aβ, tau, GFAP, and neurofilament light (NfL)—they are often invasive, costly, or lack sufficient predictive accuracy in early disease stages [16, 17]. This highlights the urgent need for accurate, cost‐effective, and minimally invasive biomarkers capable of predicting AD onset well in advance [13, 15, 17]. Emerging data indicate that early AD pathology provokes neuroinflammatory and oxidative stress responses, which in turn may induce conformational changes in P53 and overexpression of its unfolded isoform, U‐p53, in both central and peripheral tissues [58, 59, 60]. These molecular mechanisms provide a biological rationale for U‐p53AZ as a candidate biomarker of early AD‐related neurodegeneration. Evaluating the diagnostic performance of U‐p53AZ against established AD biomarkers is essential. While previous studies have suggested that U‐p53AZ may predict AD progression prior to clinical onset [26, 60], our findings extend this literature by showing that U‐p53AZ may also reflect white matter microstructural integrity changes—a domain less explored with conventional markers. However, this study did not directly compare U‐p53AZ to Aβ or tau markers in a head‐to‐head design; thus, any claims regarding its superiority remain speculative. Rigorous comparative studies are necessary to clarify the relative utility of U‐p53AZ. These results are also aligned with earlier studies by Piccirella et al. [61], who first demonstrated the diagnostic and prognostic value of U‐p53AZ in differentiating AD and non‐AD participants and predicting progression years before clinical onset [62]. Their work reported high area under curve (AUC) values (over 90% in CN and 96% in MCI), suggesting that the test could potentially outperform traditional imaging modalities such as amyloid‐PET in terms of diagnostic accuracy and prognostic power. Our results build on this by demonstrating novel associations between U‐p53AZ and DTI‐derived markers of white matter integrity—a relationship not previously reported. This study has several limitations. The modest sample size and cross‐sectional design limit the ability to infer causality or generalize the findings. The absence of a clinically diagnosed AD dementia group and the lack of MCI subtyping (e.g., amnestic vs. non‐amnestic) also constrain interpretability. While the clinical implications of our findings may be significant, they must be interpreted with caution given the exploratory nature of the analysis. Future research should involve larger, longitudinal cohorts with diverse clinical phenotypes to evaluate the temporal evolution of U‐p53AZ, its interplay with other biomarkers, and its prognostic relevance. If validated, U‐p53AZ could serve as a complementary biomarker alongside current modalities, particularly in settings where CSF collection or PET imaging is not feasible. Conclusion Our findings suggest that U‐p53AZ, as measured by the AlzoSure® Predict blood assay, may serve as a promising non‐invasive biomarker reflecting early white matter microstructural alterations, especially in individuals with mild cognitive impairment. The observed associations, particularly in the cingulum hippocampal bundle, overlap with known AD‐related neurodegenerative pathways and mirror patterns seen with CSF tau biomarkers. While these results did not withstand correction for multiple comparisons, they underscore the potential of U‐p53AZ for identifying early neurobiological changes in the Alzheimer's disease continuum. Validation in larger, longitudinal cohorts is essential to establish its prognostic and diagnostic value in clinical practice. Ethics Statement The study was conducted in accordance with the Declaration of Helsinki. All ADNI data used were de‐identified, and the use of these data for research purposes was approved by the respective institutional review boards. Ethical approval and data use permission were obtained from ADNI and relevant bodies before analysis. Author Contributions Hamide Nasiri: formal analysis, data curation. Setayesh Farahani: supervision, writing – review and editing. Mehrdad Mozafar: writing – review and editing; writing – original draft. Yasaman Ghiasi: writing – original draft, investigation. Atousa Moghadam Fard: writing – review and editing, project administration. Nasim Naseri: writing – original draft, investigation. Amirhossein Khodadadi: writing – original draft, methodology. Ali Salimi Asl: writing – original draft, methodology. Maryam Mirzapour: formal analysis; data curation; software. Hossein Hamidi: writing – original draft, investigation. Fatemeh Jafarirazi: data curation, formal analysis, software. Navid Dinarvand: software, formal analysis, data curation. Funding This research did not receive any specific grant from funding agencies in the public, commercial, or not‐for‐profit sectors. The supporting sources had no role in study design; collection, analysis, and interpretation of data; writing of the report; or the decision to submit the article for publication. Conflicts of Interest The authors declare no conflicts of interest. Acknowledgments Data collection and sharing for this project was funded by the Alzheimer's Disease Neuroimaging Initiative (ADNI) (National Institutes of Health Grant U01 AG024904) and DOD ADNI (Department of Defense award number W81XWH‐12‐2‐0012). ADNI is funded by the National Institute on Aging, the National Institute of Biomedical Imaging and Bioengineering, and through generous contributions from the following: AbbVie, Alzheimer's Association; Alzheimer's Drug Discovery Foundation; Araclon Biotech; BioClinica Inc.; Biogen; Bristol‐Myers Squibb Company; CereSpir Inc.; Cogstate; Eisai Inc.; Elan Pharmaceuticals Inc.; Eli Lilly and Company; EuroImmun; F. Hoffmann‐La Roche Ltd and its affiliated company Genentech Inc.; Fujirebio; GE Healthcare; IXICO Ltd.; Janssen Alzheimer Immunotherapy Research & Development LLC.; Johnson & Johnson Pharmaceutical Research & Development LLC.; Lumosity; Lundbeck; Merck & Co. Inc.; Meso Scale Diagnostics LLC.; NeuroRx Research; Neurotrack Technologies; Novartis Pharmaceuticals Corporation; Pfizer Inc.; Piramal Imaging; Servier; Takeda Pharmaceutical Company; and Transition Therapeutics. The Canadian Institutes of Health Research is providing funds to support ADNI clinical sites in Canada. Private sector contributions are facilitated by the Foundation for the National Institutes of Health (www.fnih.org). The grantee organization is the Northern California Institute for Research and Education, and the study is coordinated by the Alzheimer's Therapeutic Research Institute at the University of Southern California. ADNI data are disseminated by the Laboratory for Neuroimaging at the University of Southern California. All individuals named in the acknowledgments have given permission to be included, in accordance with ICMJE recommendations. Nasiri H., Farahani S., Mozafar M., et al., “Association Between U‐p53AZ Blood Biomarker and White Matter Microstructure in Early Alzheimer's Disease: A Cross‐Sectional Study Using AlzoSure® Predict,” Health Science Reports 8 (2025): 1‐10. 10.1002/hsr2.71671. Data Availability Statement Data available on request due to privacy/ethical restrictions. The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions. References Associated Data Data Availability Statement Data available on request due to privacy/ethical restrictions. The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.