Skip to main content
NCBI home page
Human Brain Mapping logoLink to Human Brain Mapping
. 2026 Jan 2;47(1):e70447. doi: 10.1002/hbm.70447

Functional Impact Score of Mitochondrial Variants and Its Relationship With Functional Connectivity of the Brain: Potential Origins of Premature Aging in Young Adulthood

Klara Mareckova 1,2,✉, Ana Paula Mendes‐Silva 3,4,5, Radek Mareček 1, Tomáš Jordánek 1,2, Anna Pačínková 1, Jana Klánová 6, Vanessa F Gonçalves 3,4,7, Yuliya S Nikolova 3,7,✉
PMCID: PMC12757729  PMID: 41479401

ABSTRACT

Alterations in mitochondrial DNA (mtDNA) have been associated with worse cognitive abilities in older adults and premature epigenetic aging in young adulthood. However, it is not clear how mitochondrial dysfunction affects brain function in young adulthood and whether cognition‐related networks might be most affected. We tested whether mtDNA functional impact (FI) score might map onto specific patterns of between‐network functional connectivity in young adults from the European Longitudinal Study of Pregnancy and Childhood (ELSPAC). We also tested whether these relationships might be mediated by accelerated epigenetic aging, calculated using Horvath's epigenetic clock, CheekAge clock, and AltumAge clock. General connectivity method was used as a reliable marker of individual differences in brain function. We showed that a greater mtDNA FI score was associated with lower connectivity between the dorsal attention and language networks (beta = −0.41, p = 0.0007, AdjR 2 = 0.15) and that there was suggestive evidence that this relationship might be mediated by accelerated epigenetic aging calculated using Horvath's epigenetic clock in young adulthood (ab = −0.061, SE = 0.04, 95% CI [−0.163; 0.001], 90% CI [−0.142; −0.002]). These findings were independent of sex, current BMI, and current substance use. Overall, we conclude that individuals with a greater mtDNA FI score might be at greater risk of experiencing worse attention to relevant linguistic inputs, greater difficulties with speech comprehension, and verbal working memory.

Keywords: dorsal attention network, functional impact score of mitochondrial variants (mtDNA FI score), general connectivity, language network, mitochondria dysfunction, young adulthood

Key Points

  • Mitochondrial DNA (mtDNA) functional impact (FI) score predicted connectivity between dorsal attention and language networks.

  • Greater mtDNA FI score was associated with lower connectivity between the networks.

  • Mitochondrial dysfunction may lead to worse language processing and verbal memory.

Greater mitochondrial DNA functional impact score was associated with lower connectivity between dorsal attention and language networks, suggesting that mitochondrial dysfunction may lead to worse language processing and verbal memory.

graphic file with name HBM-47-e70447-g002.jpg

1. Introduction

Mutations in mitochondrial DNA (mtDNA) become increasingly common with aging (López‐Otín et al. 2013) and have been associated with age‐related neurodegenerative disorders (Keogh and Chinnery 2015). Moreover, recent research from our group revealed strong relationships between greater functional impact (FI) score of mtDNA variants and premature epigenetic as well as biological aging in young adulthood (Mareckova, Mendes‐Silva, et al. 2025). Alterations in mtDNA, including deletions, have also been associated with altered cognitive abilities in older adults (Mendes‐Silva et al. 2024; Inczedy‐Farkas et al. 2014; Picard and McEwen 2014).

Mitochondrial dysfunction dramatically decreases energy supply to neurons as well as glucose metabolism in the brain, which may lead to neurodegeneration (Han et al. 2021). Imaging studies (Deery et al. 2023; Goyal et al. 2017; Subtirelu et al. 2023) consistently show that global cerebral metabolic activity decreases with age, primarily in the gray matter regions of the frontal and temporal lobes. Therefore, glucose hypometabolism in cognitively normal patients has been proposed as a predictive marker of risk for neurodegenerative disorders (Mosconi et al. 2008; Nugent et al. 2014). Brain hubs characterized by higher connectivity and thus higher energy demands are most vulnerable to deficits in energy delivery and utilization, and might be the primary targets for aging and neurodegenerative diseases (Tomasi et al. 2013). Further, a decrease in glucose consumption is closely related to a decrease in the functional activity and connectivity of the brain (Subtirelu et al. 2023; Tomasi et al. 2013). Energy consumption is also greater for longer‐range between‐network connectivity versus local within‐network connectivity (Deng et al. 2022; Takagi 2025). Thus, a greater functional impact score of mtDNA variants, compromising mitochondrial function, might particularly affect synchronization of cognition‐related networks recruiting frontal and temporal regions, with other more distal networks. If so, these between‐network connections might be targeted to prevent premature aging and the associated neurodegenerative disease.

Further research (Tomasi and Volkow 2012) aimed to test whether the age‐related decline of the brain hubs might vary across brain networks and thus might, for example, explain the age‐related decrease in attention and memory (Craik and Salthouse 2000) but relative preservation of decision making (Sanfey and Hastie 2000). Using data from the 1000 Functional Connectomes Project, they showed that aging was more strongly related to long‐range than short‐range functional connectivity and that the age‐related decreases in long‐range functional connectivity were most pronounced in the default mode and dorsal attention network (Tomasi and Volkow 2012). Given the strong relationships between the functional impact score of mtDNA variants and epigenetic as well as biological aging in young adulthood (Mareckova, Mendes‐Silva, et al. 2025) and the high energetic demands of between‐network connectivity, the FI score of mtDNA variants might map onto specific patterns of between‐network functional connectivity of the age‐sensitive brain networks in young adulthood.

Recent research on the functional connectivity of the brain introduced the general connectivity method (GFC) as a reliable marker of individual differences in brain function and cognitive abilities (Elliott et al. 2019). They showed that many fMRI studies do not collect enough resting‐state data to generate reliable measures of functional connectivity necessary for studying individual differences (Elliott et al. 2019). Therefore, the GFC method leverages shared features across resting state and task fMRI and, thanks to the length of the combined scan time, offers better test–retest reliability and thus better ability to identify meaningful correlates of individual differences in behavior (Elliott et al. 2019). The concept of the GFC method is in agreement with previous research, which has demonstrated that functional networks extracted from task fMRI are similar to those extracted from resting state fMRI (Arfanakis et al. 2000; Fair et al. 2007; Fox et al. 2006; Smith et al. 2009). According to Elliott et al. (2019), the combination of task and resting‐state fMRI increases reliability, replicability, and power and thus significantly improves the neuroscientific research on individual differences.

Building on the GFC method, within the current study, we tested the potential relationship between the mtDNA FI score, recently developed by our group (Mareckova, Mendes‐Silva, et al. 2025), and the general functional connectivity of the brain. Since examining relationships between networks across the whole brain is important in understanding overall brain functioning (Zhang et al. 2021), we took a whole‐brain approach and studied the relationships between the mtDNA FI score and GFC between the 12 main functional networks (Ji et al. 2019), namely the primary visual, secondary visual, somatomotor, cingular‐opercular, dorsal attention, language, frontoparietal, auditory, default mode, posterior multimodal, ventral multimodal, and orbito‐affective networks. This partition of brain networks was chosen because it better reflects the neurobiological definition of cortical network organization than the widely used but statistically principled partition into seven networks (Yeo et al. 2011), which merges the somatomotor and auditory systems. Based on the literature reviewed above, we hypothesized that a greater mtDNA FI score (and thus a greater presence of non‐synonymous common mutations in mtDNA with evidence for pathological outcomes) will be associated with lower GFC between the age‐sensitive brain networks. We also hypothesized that these relationships between the mtDNA FI score and lower between‐network GFC in these age‐sensitive networks might be mediated by accelerated epigenetic aging in young adulthood.

2. Materials and Methods

2.1. Participants

Participants were members of the European Longitudinal Study of Pregnancy and Childhood (ELSPAC) (Piler et al. 2017), a prenatal birth cohort born between 1991 and 1992 in South Moravia, Czech Republic, who also took part in its two follow‐ups in young adulthood—(1) Biomarkers and Underlying mechanisms of Vulnerability to Depression (VULDE; age 23–24) and (2) Health Brain Age (HBA; age 28–30) study at the Central European Institute of Technology, Masaryk University. A total of 71 young adults (all White Caucasians, mean age 29.18 years, SD = 0.50, 49.3% women) had good‐quality genetic data from the VULDE study as well as good‐quality functional magnetic resonance imaging (MRI) data (resting state as well as two fMRI tasks, totaling 40 min of fMRI data) from the HBA study and thus could be included in the current research. All participants provided written informed consent to participate in the VULDE and HBA studies, including the agreement to merge data from VULDE, HBA, and their historical data from ELSPAC. Ethical approval for the VULDE and HBA studies was obtained from the ELSPAC ethics committee (ELSPAC/EK/2/2020), and all methods were performed in accordance with the relevant guidelines and regulations.

2.2. Analysis of Genetic Data and Calculation of the mtDNA FI Score

Within the VULDE study in the early 20s, participants provided buccal swabs, DNA was isolated, and genetic analyses were performed as detailed in our previous work (Mareckova, Mendes‐Silva, et al. 2025). Briefly, we genotyped a set of 201 mitochondrial SNPs in a single batch using Illumina OmniExpressExome BeadArray 8 version 1.4. Samples that passed autosomal quality control (QC) were selected for further analysis (n = 102). Next, mtDNA was validated against the revised Cambridge Reference Sequence (rCRS) to ensure proper mapping. Following QC, we filtered out variants with a genotyping rate of 5%, individuals exhibiting missing data rates exceeding 5%, and mtDNA contamination. The total genotyping rate in the remaining samples exceeded 99%, with 198 variants and 99 individuals remaining. Of these, 20 coding mtDNA variants with a minor allele frequency (MAF) greater than 5% and 178 with MAF ≤ 5%.

mtDNA imputation was conducted as described in (Gonçalves et al. 2018), performed using IMPUTE2 v.2 software (Howie et al. 2012) and a reference panel comprising 7141 public European mitochondrial sequences obtained from the Human Mitochondrial Database (Ingman and Gyllensten 2006), adding 11 common SNPs (considering post‐imputation filters of “info” score > 0.3 and MAF > 5%), giving us 31 variants in total.

As previously described in (Mareckova, Mendes‐Silva, et al. 2025), we assessed the impact of amino acid changes through a combination of tools, MutPred (Li et al. 2009), mtDNA Selection (Pereira et al. 2011), and MitoTool (Fan and Yao 2011), that utilize sequence homology, evolutionary conservation, and protein structural information (Dong et al. 2015). For each individual, a functional impact (FI) score was calculated by summing the predictions from all three tools for the non‐synonymous common variants, with higher s indicating a greater likelihood of pathogenicity.

The FI score was calculated based on 7 potentially pathogenic mtDNA variants, namely m.4917A > G and m.5460G > A in the gene MT‐ND2, m.9477G > A in the gene MT‐CO3, m.10398A > G in the gene MT‐ND3, m.13708G > A in the gene MT‐ND5, and m.14798T > C and m.15452C > A in the gene MT‐CYB.

2.3. MRI Acquisition

MRI of the brain was conducted within the HBA study in the late 20s using a 3 T Siemens Prisma MRI scanner. T1‐weighted (T1w) whole‐brain MPRAGE images were acquired using a 64‐channel head/neck coil with acquisition parameters: voxel size 1 mm3, repetition time (TR) 2300 ms, echo time (TE) 2.34 ms, inversion time (TI) 900 ms, and flip angle 8°. During the same session, functional MRI images were acquired during rest and during two fMRI tasks: (1) emotion regulation task with negative and neutral images from the International Affective Picture System (IAPS) and (2) Monetary Incentive Delay (MID) task. A detailed description of these two fMRI tasks is provided in Mareckova, Trbusek, et al. (2025) and Mareckova et al. (2024), respectively.

The IAPS and MID tasks were acquired using the same multi‐echo, multi‐band echo planar imaging (EPI) sequence with TR = 700 ms, TE = 0.0164 s (Echo 1), 0.03766 s (Echo 2), 0.05892 s (Echo 3), a multi‐band factor of 6, FOV = 210 × 210 mm, in‐plane resolution 70 × 70, 48 transversal slices with thickness of 3 mm, flip angle 47°. The sequences for IAPS and MID tasks had 2010 and 935 volumes, respectively. We also acquired a gradient echo field map with the same geometric settings, two echo times (4.92 and 7.38 ms), TR = 520 ms and flip angle of 60°. The RS was acquired using single‐echo, single‐band EPI with TR = 2.08 s, TE = 30 ms, FOV = 192 × 192 mm, in‐plane resolution 64 × 64, 39 transversal slices with thickness of 3 mm, flip angle 90°, 200 volumes. The RS sequence was set to match previous studies whose data are not used in this study.

2.4. Pre‐Processing of MRI Data

We performed the preprocessing of fMRI data using the SPM12 toolbox run under MATLAB R2017b. The data for the second echo was realigned to the first volume, and the same transformation was applied also to the first and third echo data. The estimated voxel displacement map (VDM) was then applied to echo data to correct for geometry distortions. Data from the three echoes were fused in a voxel‐wise manner by the weighted sum according to the temporal‐signal‐to‐noise ratio (Murphy et al. 2007). In the case of the RS data, the echo fusion and VDM correction were skipped as we had single‐echo data and did not acquire a field map.

Next, the structural scan was coregistered to the mean functional volume and normalized to the MNI space. The resulting transformation was applied to the functional volumes. To facilitate data fusion from the three datasets, the timeseries for each voxel was normalized to an average value of 100. To correct for high amplitude spikes in the data, we performed despiking using BrainWavelet toolbox and the function wdscore. To perform correction for physiological noise caused by the subject's movement, respiration, and blood pulsation, we applied regression‐based filtering in the time domain. The set of regressors consisted of low‐frequency harmonic signals up to the frequency of 1/125 Hz, movement parameters, and their differences coming from the realign step, the first five principal components of signals from white matter (WM) and cerebrospinal fluid (CSF, according to MNI template masks of WM and CSF), and the global signal estimated as an average of all in‐brain voxels. In the case of the two tasks, the set of regressors further contained additional ones that modeled a response to stimulation as outlined below.

In the case of IAPS, the task was modeled by 10 regressors: initial part of a trial (3 s), active part of a trial (5 s) and the response part of a trial (3 s) for each of the three conditions (neutral, observe negative, and regulate negative) and another one for modeling the subject's responses by key‐strokes. In the case of MID, the task was modeled by 12 regressors: cue and anticipation parts of a trial for each condition (rewards, losses, and neutral), a regressor modeling a probe part of a trial, and then five feedback regressors (gain or no gain in reward trial, lose or not lose in loss trial and feedback in neutral trial). All task regressors were constructed as a convolution of respective boxcar function or events with a canonical hemodynamic response function. Finally, the data were smoothed by a 6 mm Gaussian spatial filter. The amount of subjects' movement during the data acquisition was checked by the average Frame‐wise Displacement metric (FD). All subjects that had FD > 0.3 mm in any dataset were discarded from the subsequent analyses, as recommended by previous research (Power et al. 2014).

The pre‐processed data were parcellated according to the Glasser atlas (Glasser et al. 2016). For each ROI, the representative time‐series were computed by averaging the time‐series of voxels belonging to the ROI.

2.5. Connectivity Analysis

Preprocessed and parcellated data from two task measurements and resting state sessions were concatenated together. As the length of the measurements was not the same for all the subjects, the data from each session was cropped to the same length as follows: 2010 scans in the first task, 935 scans in the second task, and 200 scans in the resting state data. The concatenated data, parcellated into 360 regions, were merged into 12 functional networks (Ji et al. 2019) as a mean of all belonging regions. Next, between‐network static functional connectivity was calculated between each pair of the 12 functional networks as their Pearson's correlation, and Fischer's z‐transformation was applied. This resulted in a 12 × 12 connectivity matrix with zeroes on the diagonal, as the within‐network connectivity was not calculated. The matrix is symmetrical along the diagonal, so the upper triangle is used only.

2.6. Epigenetic Aging in Young Adulthood

Buccal swabs were collected from all participants of the HBA study in the late 20s. DNA methylation was assessed with the Illumina EPIC Platform. Epigenetic age was calculated using three different epigenetic clocks, namely the Horvath's epigenetic clock (Horvath 2013), which we have used also in our previous work (Mareckova et al. 2023; Marečková et al. 2020), the AltumAge clock (De Lima Camillo et al. 2022), and the CheekAge clock (Shokhirev et al. 2024).

Briefly, to process the raw Illumina microarray data, the R package ChAMP was used (Tian et al. 2017). Beta mixture quantile normalization (BMIQ) (Teschendorff et al. 2013) method was used to adjust the beta‐values of type II design probes into a statistical distribution characteristic of type I probes. DNA methylation age was calculated according to Horvath's approach (Horvath 2013) using 353 CpG sites for the estimation of epigenetic age, according to the AltumAge clock (De Lima Camillo et al. 2022) using 20,318 CpG sites, and according to the CheekAge clock (Shokhirev et al. 2024) using 115,533 CpG sites. See Methods S1 for details and the link to GitHub Repository. For each of these clocks, the epigenetic age gap estimate (EpiAGE) was calculated as the DNA methylation age residualized for batch, chronological age, and the proportion of epithelial cells (the average proportion was approximately 78% of epithelial and 22% of immune cells; SD = 22% in each participant). Positive values of EpiAGE reflect accelerated aging, and negative values reflect decelerated aging.

2.7. Covariates

Information regarding substance use in young adulthood was collected within the HBA study in the late 20s using a self‐reported questionnaire. This questionnaire asked (1) How many times in the past 30 days did you smoke cigarettes, (2) How many times in the past 30 days did you drink a beer, wine, or shots, and (3) How many times in your life did you use cannabis?

2.8. Statistical Analysis

Statistical analysis was done in MATLAB (version: 23.2.0.2365128 [R2023b], Natick, Massachusetts: MathWorks Inc.). First, we assessed the distribution of data, and variables that did not follow a normal distribution (e.g., FI score of mtDNA variants) were transformed using logarithmic transformation. Next, we used linear regression to assess the impact of the FI score of mtDNA variants on GFC in the main functional networks, and multiple comparisons were corrected using the False Discovery Rate (FDR) method. Similar to our previous work (Mareckova, Mendes‐Silva, et al. 2025), the covariates included sex, current BMI, cigarette smoking, cannabis, and alcohol use.

Finally, a mediation analysis assessed whether the relationship between the mtDNA FI score and functional connectivity between the brain networks associated with mtDNA FI score might be mediated by accelerated epigenetic aging, as assessed by Horvath's clock, CheekAge clock, and AltumAge clock in young adulthood. For consistency with the linear regression analyses above, sex, current BMI, cigarette smoking, cannabis, and alcohol use were used as covariates. Model 4 of the PROCESS macro (Hayes 2018) for SPSS, version 27 (IBM SPSS Statistics, IBM Corp, Armonk NY) was used to compute the indirect effects. Conditional indirect effects were obtained for significance using bootstrapped (10,000 iterations) bias‐corrected 95% confidence intervals constructed around indirect effect estimates.

3. Results

3.1. Relationship Between the FI Score of mtDNA Variants and GFC of the Brain

The FI score of mtDNA variants was significantly linked to the connectivity between the dorsal attention and language network (F = 12.79, p = 0.0007, FDRp = 0.0440; see overview of results in Table 1 and detailed statistics for all 66 models in Supporting Information). Since there was no significant interaction with sex, the final model used mtDNA FI score as the main predictor and sex, current BMI, current cigarette smoking, alcohol use, and lifetime cannabis use as covariates. The results remained significant even after these adjustments and subsequent FDR correction for the number of comparisons (see Table 2). Post hoc analyses revealed that a greater FI score of mtDNA variants was significantly associated with lower connectivity between the dorsal attention and language network (beta = −0.41, p = 0.0007, AdjR 2 = 0.15; Figure 1). The results remained significant even when excluding the one potential outlier (beta = −0.34, p = 0.004, AdjR 2 = 0.15).

TABLE 1.

MtDNA FI score and general connectivity.

VIS1 VIS2 SOM CIN‐OP DOR‐ATT LAN FP AUD DMN POST‐MU VEN‐MUL ORB‐AFF
VIS1 0.000 2.803 2.027 0.927 7.489 6.764 1.112 0.022 2.253 1.127 0.128 1.970
VIS2 2.803 0.000 1.477 2.056 6.418 4.031 0.336 0.014 2.179 1.615 0.996 1.235
SOM 2.027 1.477 0.000 0.561 0.064 0.017 1.974 1.687 1.968 2.103 1.670 2.632
CIN‐OP 0.927 2.056 0.561 0.000 5.335 0.384 0.926 0.074 1.908 0.269 2.148 4.775
DOR‐ATT 7.489 6.418 0.064 5.335 0.000 12.799 3.632 2.426 6.012 7.880 6.387 2.035
LAN 6.764 4.031 0.017 0.384 12.799 0.000 0.739 2.034 0.792 3.019 0.934 2.657
FP 1.112 0.336 1.974 0.926 3.632 0.739 0.000 0.054 0.052 0.757 0.553 0.238
AUD 0.022 0.014 1.687 0.074 2.426 2.034 0.054 0.000 0.017 0.514 0.024 0.663
DMN 2.253 2.179 1.968 1.908 6.012 0.792 0.052 0.017 0.000 0.145 0.011 0.525
POST‐MU 1.127 1.615 2.103 0.269 7.880 3.019 0.757 0.514 0.145 0.000 0.832 2.071
VEN‐MUL 0.128 0.996 1.670 2.148 6.387 0.934 0.553 0.024 0.011 0.832 0.000 1.095
ORB‐AFF 1.970 1.235 2.632 4.775 2.035 2.657 0.238 0.663 0.525 2.071 1.095 0.000
Open in a new tab

Note: mtDNA FI score was associated with general connectivity between dorsal attention and the language network. The numbers are F‐statistics, and the shade of red color indicates the magnitude of the F‐statistics. The uncorrected significant results are in bold, and the FDR‐corrected significant results are in bold square.

TABLE 2.

Regression table revealing the significant effect of the FI score of mtDNA variants on connectivity between the dorsal attention and language network (F = 12.79, p = 0.0007, FDRp = 0.0440), which was independent of sex, current BMI, current cigarette smoking, alcohol use, and lifetime cannabis use.

Dorsal‐attention ↔ Language
F df p AdjR 2
Model statistics 3.048 64 0.011 0.136
Regressor Beta SE F p
Regressor statistics
Sex 0.017 0.066 0.065 0.800
FI score −0.411 0.115 12.799 0.001
BMI −0.288 0.199 2.095 0.153
Smoking −0.016 0.022 0.511 0.477
Cannabis use 0.010 0.013 0.601 0.441
Alcohol use −0.008 0.008 0.993 0.323
Open in a new tab

FIGURE 1.

FIGURE 1

Open in a new tab

A Greater mtDNA FI score was associated with lower connectivity between dorsal attention and the language network (beta = −0.41, p = 0.0007, AdjR 2 = 0.15), and these relationships were independent of sex, current BMI, current cigarette smoking, alcohol use, and lifetime cannabis use. The red dashed lines represent the confidence interval, and the blue dashed line represents the mean.

A greater FI score of mtDNA variants was also associated with lower between‐network connectivity of the dorsal attention (DOR‐ATT) network with visual (VIS1: F = 7.48, p = 0.0080, FDRp = 0.1537; VIS2: F = 6.42, p = 0.0138, FDRp = 0.1537), cingulo‐opercular (CIN‐OP: F = 5.34, p = 0.0241, FDRp = 0.1991), default mode (DMN: F = 6.01, p = 0.0169, FDRp = 0.1598), posterior multimodal (POST‐MU: F = 7.89, p = 0.0066, FDRp = 0.1537), and ventral multimodal (VEN‐MUL: F = 6.39, p = 0.0140, FDRp = 0.1537) networks, as well as lower connectivity between the language and visual (VIS1: F = 6.76, p = 0.0115, FDRp = 0.1537) networks. However, none of these relationships survived the FDR correction (see Table 1 for the overview and Table S3 for all statistical details).

3.2. Does Accelerated Epigenetic Aging in Young Adulthood Mediate the Relationship Between the Greater mtDNA FI Score and Functional Connectivity Between Dorsal Attention and Language Network?

The mediation of the relationship between higher mtDNA FI score and lower functional connectivity between dorsal attention and language network by accelerated epigenetic aging did not reach significance when using the 95% confidence interval (Horvath's clock: ab = −0.061, SE = 0.04, 95% CI [−0.163; 0.001]; CheekAge clock: ab = −0.007, SE = 0.022, 95% CI [−0.065; 0.024]; AltumAge clock: ab = −0.007, SE = 0.036, 95% CI [−0.089; 0.059]). However, suggestive evidence of such mediation by Horvath's clock appeared when using a 90% confidence interval (ab = −0.061, SE = 0.04, 90% CI [−0.142; −0.002]; Figure S1), and these relationships were independent of sex, current BMI, current cigarette smoking, alcohol use, and lifetime cannabis use.

4. Discussion

We studied the FI of mtDNA variants on the GFC of the brain in young adulthood. Our findings showed that a greater FI score of mtDNA variants, which is based on seven non‐synonymous common variants and has been previously associated with accelerated epigenetic and biological aging in young adulthood (Mareckova, Mendes‐Silva, et al. 2025), was associated with lower GFC between the dorsal attention and language networks in young adulthood. These results survived the FDR correction as well as the correction for covariates, including sex, current BMI, and current substance use. Moreover, we provided suggestive evidence that the relationship between a greater mtDNA FI score and lower connectivity between the dorsal attention and language networks might be mediated by accelerated epigenetic aging calculated using Horvath's epigenetic clock in young adulthood.

The dorsal attention network is essential for holding attention and thus contributes to a number of cognitive skills. It includes the dorsal frontoparietal regions, such as the intraparietal sulcus and frontal eye fields, and specializes in the voluntary deployment of attention (Vossel et al. 2014). The language network includes the frontal and temporal areas and is, in most individuals, lateralized in the left hemisphere (Fedorenko et al. 2024). It is involved in storing language knowledge and accessing words (Fedorenko et al. 2024). The lower connectivity between these two networks might thus reflect lower orienting and maintaining attention toward speech or text and thus lead to a worse understanding and worse cognitive skills, which are often reported in people with age‐related disorders such as mild cognitive impairment or neurodegenerative disease. Our findings thus suggest that individuals with a greater mtDNA FI score might be at greater risk of experiencing worse attention to relevant linguistic inputs, and have a greater risk of worse language processing, speech comprehension, and verbal working memory. We also speculate that the lower connectivity between dorsal attention and the language network might reflect difficulties in filtering out distractions during verbal working memory or other language tasks, as well as worse attention to conversations.

These findings substantially extend our previous research on mtDNA FI score and accelerated aging. Moreover, they also substantially extend previous research in mice, which reported that alterations in mtDNA were associated with alterations of cerebral beta‐amyloid accumulation, suggesting the development of age‐related diseases such as Alzheimer disease (Scheffler et al. 2012). The current findings demonstrate that alterations in mtDNA and the associated changes in functional connectivity between the dorsal attention and language networks are detectable already in healthy young adults. This further suggests that lower functional connectivity between these networks might serve as an early marker of higher risk for age‐related disorders.

Pathogenic mutations of mtDNA result in impaired mitochondrial oxidative phosphorylation (OXPHOS), increased generation of reactive oxygen species (ROS), and reduced energy production (Scheffler et al. 2012). The specific mutations in the MT‐ND2, MT‐ND3, MT‐ND4, MT‐CO3, and MT‐CYB genes, which characterize the mtDNA FI score used in the current study and which are known to have functional effects related to OXPHOS complexes I, II, and IV, have been previously associated with Alzheimer's disease (Lunnon et al. 2017; Manczak et al. 2004), Parkinson's disease (Vos 2022), bipolar disorder (Tachi et al. 2023; Kato et al. 2001, 2000; McMahon et al. 2000) and type 2 diabetes mellitus (Prasun 2020; Soini et al. 2013). It might be that these associations, as well as the associations between the mtDNA FI score and lower connectivity between the dorsal attention and language network reported in the current study, might be mediated by accelerated epigenetic aging. The suggestive evidence for the latter mediation provided in the current study using Horvath's epigenetic clock fits with previous research that has reported that disruptions or mutations in mtDNA can result in epigenetic changes in nuclear DNA (Arfanakis et al. 2000; Bellizzi et al. 2012; Lopes 2020). As explained in a review on mitochondria and DNA methylation (Lopes 2020), mitochondria control nuclear DNA methylation via one‐carbon metabolism, also known as the folate cycle, tricarboxylic acid cycle, and methionine pathway. However, we did not find support for these relationships using AltumAge or CheekAge epigenetic clocks, and thus, further research in cohorts containing blood‐based epigenetic data is needed to resolve this inconsistency and provide a better understanding of the exact mechanisms between mtDNA FI score, DNA methylation, and brain function.

Our findings are also limited by the small sample size, which may restrict the generalizability of our results to broader populations. Additionally, the cross‐sectional nature of our study prevents us from establishing any causal relationships between variants and brain connectivity. Validation in larger cohorts, longitudinal study designs, and the inclusion of diverse ancestry populations are essential to generate comprehensive insights. Still, the combination of data on mtDNA variants, DNA methylation, and functional connectivity of the brain, and particularly the use of the novel FI score of mtDNA variants, as well as the method of GFC are clear strengths offering a unique insight into the developmental origins of premature aging.

Author Contributions

K.M. conceptualization, formal analysis, writing – original draft. A.P.M.‐S. methodology and mtDNA analyses. R.M. methodology and pre‐processing of fMRI data. T.J. methodology and general connectivity analyses. A.P. methodology and epigenetic analyses. J.K. resources. V.F.G. supervision of mtDNA analyses. Y.S.N. conceptualization, writing – review and editing.

Funding

This study was supported by Grantová Agentura České Republiky (24‐12183M), Agentura Pro Zdravotnický Výzkum České Republiky (NU20J‐04‐00022), European Union—Next Generation EU (LX22NPO5107), Ministerstvo Školství, Mládeže a Tělovýchovy (LM2018129 and LM2023069), Horizon 2020 Framework Programme (857560), Centre for Addiction and Mental Health Foundation, and Ministry of Education, Youth and Sports, Czech Republic (CEITEC 2020 and LQ1601).

Ethics Statement

This study was approved by the ELSPAC Ethics Committee, and all participants signed informed consent.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Data S1: Supporting information.

HBM-47-e70447-s001.docx (112.6KB, docx)

Acknowledgments

This work has received funding from the Czech Science Foundation, project no. 24‐12183M, Czech Health Research Council (No. NU20J‐04‐00022), the Czech Ministry of Education, Youth and Sports (MEYS CR) (CEITEC 2020, LQ1601), and by project no. LX22NPO5107 (MEYS): Funded by European Union—Next Generation EU. Support with obtaining scientific data presented in this paper came from the core facility Multimodal and Functional Imaging Laboratory of Central European Institute of Technology, Masaryk University, supported by the Czech‐BioImaging large RI project (No. LM2018129, funded by MEYS CR). Authors also thank the RECETOX Research Infrastructure (No. LM2023069), financed by MEYS CR for its supportive background. This work was also supported by the European Union's Horizon 2020 research and innovation program under grant agreement No. 857560 (CETOCOEN Excellence). This publication reflects only the author's view, and the European Commission is not responsible for any use that may be made of the information it contains. Dr. Mendes‐Silva acknowledges support from CIHR Fellowship Award, the CAMH Discovery Fund Fellowship, and 2023 NARSAD Young Investigator Grant from Brain & Behavior Research Foundation. Dr. Gonçalves is supported by Larry and Judy Tanenbaum Foundation, as well as the CAMH Foundation Discovery Fund and WomenMind Grants. Dr. Nikolova is supported by Koerner New Scientist Award from the CAMH Foundation and a Discovery Grant from the National Sciences and Engineering Research Council of Canada (NSERC).

Mareckova, K. , Mendes‐Silva A. P., Mareček R., et al. 2026. “Functional Impact Score of Mitochondrial Variants and Its Relationship With Functional Connectivity of the Brain: Potential Origins of Premature Aging in Young Adulthood.” Human Brain Mapping 47, no. 1: e70447. 10.1002/hbm.70447.

Contributor Information

Klara Mareckova, Email: klara.mareckova@ceitec.muni.cz.

Yuliya S. Nikolova, Email: yuliya.nikolova@camh.ca.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

References

  1. Arfanakis, K. , Cordes D., Haughton V. M., Moritz C. H., Quigley M. A., and Meyerand M. E.. 2000. “Combining Independent Component Analysis and Correlation Analysis to Probe Interregional Connectivity in fMRI Task Activation Datasets.” Magnetic Resonance Imaging 18, no. 8: 921–930. 10.1016/s0730-725x(00)00190-9. [DOI] [PubMed] [Google Scholar]
  2. Bellizzi, D. , D'Aquila P., Giordano M., Montesanto A., and Passarino G.. 2012. “Global DNA Methylation Levels Are Modulated by Mitochondrial DNA Variants.” Epigenomics 4, no. 1: 17–27. 10.2217/epi.11.109. [DOI] [PubMed] [Google Scholar]
  3. Craik, F. , and Salthouse T.. 2000. The Handbook of Aging and Cognition. Lawrence Erlbaum Associates Inc. [Google Scholar]
  4. De Lima Camillo, L. P. , Lapierre L. R., and Singh R. A.. 2022. “A Pan‐Tissue DNA‐Methylation Epigenetic Clock Based on Deep Learning.” NPJ Aging 8: 4. 10.1038/s41514-022-00085-y. [DOI] [Google Scholar]
  5. Deery, H. A. , Di Paolo R., Moran C., Egan G. F., and Jamadar S. D.. 2023. “Lower Brain Glucose Metabolism in Normal Ageing Is Predominantly Frontal and Temporal: A Systematic Review and Pooled Effect Size and Activation Likelihood Estimates Meta‐Analyses.” Human Brain Mapping 44, no. 3: 1251–1277. 10.1002/hbm.26119. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Deng, S. , Li J., Thomas Yeo B. T., and Gu S.. 2022. “Control Theory Illustrates the Energy Efficiency in the Dynamic Reconfiguration of Functional Connectivity.” Communications Biology 5, no. 1: 295. 10.1038/s42003-022-03196-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Dong, C. , Wei P., Jian X., et al. 2015. “Comparison and Integration of Deleteriousness Prediction Methods for Nonsynonymous SNVs in Whole Exome Sequencing Studies.” Human Molecular Genetics 24, no. 8: 2125–2137. 10.1093/hmg/ddu733. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Elliott, M. L. , Knodt A. R., Cooke M., et al. 2019. “General Functional Connectivity: Shared Features of Resting‐State and Task fMRI Drive Reliable and Heritable Individual Differences in Functional Brain Networks.” NeuroImage 189: 516–532. 10.1016/j.neuroimage.2019.01.068. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Fair, D. A. , Dosenbach N. U., Church J. A., et al. 2007. “Development of Distinct Control Networks Through Segregation and Integration.” Proceedings of the National Academy of Sciences of the United States of America 104, no. 33: 13507–13512. 10.1073/pnas.0705843104. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Fan, L. , and Yao Y. G.. 2011. “MitoTool: A Web Server for the Analysis and Retrieval of Human Mitochondrial DNA Sequence Variations.” Mitochondrion 11, no. 2: 351–356. 10.1016/j.mito.2010.09.013. [DOI] [PubMed] [Google Scholar]
  11. Fedorenko, E. , Ivanova A. A., and Regev T. I.. 2024. “The Language Network as a Natural Kind Within the Broader Landscape of the Human Brain.” Nature Reviews. Neuroscience 25, no. 5: 289–312. 10.1038/s41583-024-00802-4. [DOI] [PubMed] [Google Scholar]
  12. Fox, M. D. , Snyder A. Z., Zacks J. M., and Raichle M. E.. 2006. “Coherent Spontaneous Activity Accounts for Trial‐to‐Trial Variability in Human Evoked Brain Responses.” Nature Neuroscience 9, no. 1: 23–25. 10.1038/nn1616. [DOI] [PubMed] [Google Scholar]
  13. Glasser, M. F. , Coalson T. S., Robinson E. C., et al. 2016. “A Multi‐Modal Parcellation of Human Cerebral Cortex.” Nature 536, no. 7615: 171–178. 10.1038/nature18933. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Gonçalves, V. F. , Giamberardino S. N., Crowley J. J., et al. 2018. “Examining the Role of Common and Rare Mitochondrial Variants in Schizophrenia.” PLoS One 13, no. 1: e0191153. 10.1371/journal.pone.0191153. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Goyal, M. S. , Vlassenko A. G., Blazey T. M., et al. 2017. “Loss of Brain Aerobic Glycolysis in Normal Human Aging.” Cell Metabolism 26, no. 2: 353–360.e353. 10.1016/j.cmet.2017.07.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Han, R. , Liang J., and Zhou B.. 2021. “Glucose Metabolic Dysfunction in Neurodegenerative Diseases‐New Mechanistic Insights and the Potential of Hypoxia as a Prospective Therapy Targeting Metabolic Reprogramming.” International Journal of Molecular Sciences 22, no. 11: 5887. 10.3390/ijms22115887. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Hayes, A. F. 2018. Introduction to Mediation, Moderation, and Conditional Process Analysis. Second Edition. A Regression‐Based Approach. Guilford Press. [Google Scholar]
  18. Horvath, S. 2013. “DNA Methylation Age of Human Tissues and Cell Types.” Genome Biology 14, no. 10: R115. 10.1186/gb-2013-14-10-r115. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Howie, B. , Fuchsberger C., Stephens M., Marchini J., and Abecasis G. R.. 2012. “Fast and Accurate Genotype Imputation in Genome‐Wide Association Studies Through Pre‐Phasing.” Nature Genetics 44, no. 8: 955–959. 10.1038/ng.2354. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Inczedy‐Farkas, G. , Trampush J. W., Perczel Forintos D., et al. 2014. “Mitochondrial DNA Mutations and Cognition: A Case‐Series Report.” Archives of Clinical Neuropsychology 29, no. 4: 315–321. 10.1093/arclin/acu016. [DOI] [PubMed] [Google Scholar]
  21. Ingman, M. , and Gyllensten U.. 2006. “ mtDB: Human Mitochondrial Genome Database, a Resource for Population Genetics and Medical Sciences.” Nucleic Acids Research 34, no. Database issue: D749–D751. 10.1093/nar/gkj010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Ji, J. L. , Spronk M., Kulkarni K., Repovš G., Anticevic A., and Cole M. W.. 2019. “Mapping the Human Brain's Cortical–Subcortical Functional Network Organization.” NeuroImage 185: 35–57. 10.1016/j.neuroimage.2018.10.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Kato, T. , Kunugi H., Nanko S., and Kato N.. 2000. “Association of Bipolar Disorder With the 5178 Polymorphism in Mitochondrial DNA.” American Journal of Medical Genetics 96: 182–186. [PubMed] [Google Scholar]
  24. Kato, T. , Kunugi H., Nanko S., and Kato N.. 2001. “Mitochondrial DNA Polymorphisms in Bipolar Disorder.” Journal of Affective Disorders 62: 151–164. [DOI] [PubMed] [Google Scholar]
  25. Keogh, M. J. , and Chinnery P. F.. 2015. “Mitochondrial DNA Mutations in Neurodegeneration.” Biochimica et Biophysica Acta 1847, no. 11: 1401–1411. 10.1016/j.bbabio.2015.05.015. [DOI] [PubMed] [Google Scholar]
  26. Li, B. , Krishnan V. G., Mort M. E., et al. 2009. “Automated Inference of Molecular Mechanisms of Disease From Amino Acid Substitutions.” Bioinformatics 25, no. 21: 2744–2750. 10.1093/bioinformatics/btp528. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Lopes, F. C. A. 2020. “Mitochondrial Metabolism and DNA Methylation: A Review of the Interaction Between Two Genomes.” Clinical Epigenetics 12, no. 1: 182. 10.1186/s13148-020-00976-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. López‐Otín, C. , Blasco M. A., Partridge L., Serrano M., and Kroemer G.. 2013. “The Hallmarks of Aging.” Cell 153, no. 6: 1194–1217. 10.1016/j.cell.2013.05.039. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Lunnon, K. , Keohane A., Pidsley R., et al. 2017. “Mitochondrial Genes Are Altered in Blood Early in Alzheimer's Disease.” Neurobiology of Aging 53: 36–47. 10.1016/j.neurobiolaging.2016.12.029. [DOI] [PubMed] [Google Scholar]
  30. Manczak, M. , Park B. S., Jung Y., and Reddy P. H.. 2004. “Differential Expression of Oxidative Phosphorylation Genes in Patients With Alzheimer's Disease: Implications for Early Mitochondrial Dysfunction and Oxidative Damage.” Neuromolecular Medicine 5, no. 2: 147–162. 10.1385/nmm:5:2:147. [DOI] [PubMed] [Google Scholar]
  31. Mareckova, K. , Marecek R., Andryskova L., Brazdil M., and Nikolova Y. S.. 2024. “Prenatal Exposure to Alcohol and Its Impact on Reward Processing and Substance Use in Adulthood.” Translational Psychiatry 14, no. 1: 220. 10.1038/s41398-024-02941-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Mareckova, K. , Mendes‐Silva A. P., Jáni M., et al. 2025. “Mitochondrial DNA Variants and Their Impact on Epigenetic and Biological Aging in Young Adulthood.” Translational Psychiatry 15, no. 1: 16. 10.1038/s41398-025-03235-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Mareckova, K. , Pacinkova A., Marecek R., et al. 2023. “Longitudinal Study of Epigenetic Aging and Its Relationship With Brain Aging and Cognitive Skills in Young Adulthood.” Frontiers in Aging Neuroscience 15: 1215957. 10.3389/fnagi.2023.1215957. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Mareckova, K. , Trbusek F., Marecek R., et al. 2025. “Maternal Depression During the Perinatal Period and Its Relationship With Emotion Regulation in Young Adulthood: An fMRI Study in a Prenatal Birth Cohort.” Psychological Medicine 55: e39. 10.1017/s0033291725000042. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Marečková, K. , Pačínková A., Klasnja A., et al. 2020. “Epigenetic Clock as a Correlate of Anxiety.” NeuroImage: Clinical 28: 102458. 10.1016/j.nicl.2020.102458. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. McMahon, F. J. , Chen Y. S., Patel S., et al. 2000. “Mitochondrial DNA Sequence Diversity in Bipolar Affective Disorder.” American Journal of Psychiatry 157: 1058–1064. [DOI] [PubMed] [Google Scholar]
  37. Mendes‐Silva, A. P. , Nikolova Y. S., Rajji T. K., et al. 2024. “Exosome‐Associated Mitochondrial DNA in Late‐Life Depression: Implications for Cognitive Decline in Older Adults.” Journal of Affective Disorders 362: 217–224. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Mosconi, L. , Pupi A., and De Leon M. J.. 2008. “Brain Glucose Hypometabolism and Oxidative Stress in Preclinical Alzheimer's Disease.” Annals of the New York Academy of Sciences 1147: 180–195. 10.1196/annals.1427.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Murphy, K. , Bodurka J., and Bandettini P. A.. 2007. “How Long to Scan? The Relationship Between fMRI Temporal Signal to Noise Ratio and Necessary Scan Duration.” NeuroImage 34, no. 2: 565–574. 10.1016/j.neuroimage.2006.09.032. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Nugent, S. , Castellano C. A., Goffaux P., et al. 2014. “Glucose Hypometabolism Is Highly Localized, but Lower Cortical Thickness and Brain Atrophy Are Widespread in Cognitively Normal Older Adults.” American Journal of Physiology. Endocrinology and Metabolism 306, no. 11: E1315–E1321. 10.1152/ajpendo.00067.2014. [DOI] [PubMed] [Google Scholar]
  41. Pereira, L. , Soares P., Radivojac P., Li B., and Samuels D. C.. 2011. “Comparing Phylogeny and the Predicted Pathogenicity of Protein Variations Reveals Equal Purifying Selection Across the Global Human mtDNA Diversity.” American Journal of Human Genetics 88, no. 4: 433–439. 10.1016/j.ajhg.2011.03.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Picard, M. , and McEwen B. S.. 2014. “Mitochondria Impact Brain Function and Cognition.” Proceedings of the National Academy of Sciences of the United States of America 111, no. 1: 7–8. 10.1073/pnas.1321881111. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Piler, P. , Kandrnal V., Kukla L., et al. 2017. “Cohort Profile: The European Longitudinal Study of Pregnancy and Childhood (ELSPAC) in The Czech Republic.” International Journal of Epidemiology 46, no. 5: 1379. 10.1093/ije/dyw091. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Power, J. D. , Mitra A., Laumann T. O., Snyder A. Z., Schlaggar B. L., and Petersen S. E.. 2014. “Methods to Detect, Characterize, and Remove Motion Artifact in Resting State fMRI .” NeuroImage 84: 320–341. 10.1016/j.neuroimage.2013.08.048. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Prasun, P. 2020. “Role of Mitochondria in Pathogenesis of Type 2 Diabetes Mellitus.” Journal of Diabetes and Metabolic Disorders 19, no. 2: 2017–2022. 10.1007/s40200-020-00679-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Sanfey, A. , and Hastie R.. 2000. Cognitive Aging: A Primer. Psychology Press. [Google Scholar]
  47. Scheffler, K. , Krohn M., Dunkelmann T., et al. 2012. “Mitochondrial DNA Polymorphisms Specifically Modify Cerebral β‐Amyloid Proteostasis.” Acta Neuropathologica 124, no. 2: 199–208. 10.1007/s00401-012-0980-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Shokhirev, M. N. , Torosin N. S., Kramer D. J., Johnson A. A., and Cuellar T. L.. 2024. “CheekAge: A Next‐Generation Buccal Epigenetic Aging Clock Associated With Lifestyle and Health.” Geroscience 46, no. 3: 3429–3443. 10.1007/s11357-024-01094-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Smith, S. M. , Fox P. T., Miller K. L., et al. 2009. “Correspondence of the Brain's Functional Architecture During Activation and Rest.” Proceedings of the National Academy of Sciences of the United States of America 106, no. 31: 13,040–13,045. 10.1073/pnas.0905267106. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Soini, H. K. , Moilanen J. S., Vilmi‐Kerälä T., Finnilä S., and Majamaa K.. 2013. “Mitochondrial DNA Variant m.15218A > G in Finnish Epilepsy Patients Who Have Maternal Relatives With Epilepsy, Sensorineural Hearing Impairment or Diabetes Mellitus.” BMC Medical Genetics 14: 73. 10.1186/1471-2350-14-73. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Subtirelu, R. C. , Teichner E. M., Su Y., et al. 2023. “Aging and Cerebral Glucose Metabolism: (18)F‐FDG‐PET/CT Reveals Distinct Global and Regional Metabolic Changes in Healthy Patients.” Life 13, no. 10: 2044. 10.3390/life13102044. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Tachi, R. , Ohi K., Nishizawa D., et al. 2023. “Mitochondrial Genetic Variants Associated With Bipolar Disorder and Schizophrenia in a Japanese Population.” International Journal of Bipolar Disorders 11: 26. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Takagi, K. 2025. “A Reduction in Energy Costs Induces Integrated States of Brain Dynamics.” Scientific Reports 15, no. 1: 11421. 10.1038/s41598-025-96120-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Teschendorff, A. E. , Marabita F., Lechner M., et al. 2013. “A Beta‐Mixture Quantile Normalization Method for Correcting Probe Design Bias in Illumina Infinium 450 k DNA Methylation Data.” Bioinformatics 29, no. 2: 189–196. 10.1093/bioinformatics/bts680. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Tian, Y. , Morris T. J., Webster A. P., et al. 2017. “ChAMP: Updated Methylation Analysis Pipeline for Illumina BeadChips.” Bioinformatics 33, no. 24: 3982–3984. 10.1093/bioinformatics/btx513. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Tomasi, D. , and Volkow N. D.. 2012. “Aging and Functional Brain Networks.” Molecular Psychiatry 17, no. 5: 471–549–558. 10.1038/mp.2011.81. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Tomasi, D. , Wang G. J., and Volkow N. D.. 2013. “Energetic Cost of Brain Functional Connectivity.” Proceedings of the National Academy of Sciences of the United States of America 110, no. 33: 13642–13647. 10.1073/pnas.1303346110. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Vos, M. 2022. “Mitochondrial Complex I Deficiency: Guilty in Parkinson's Disease.” Signal Transduction and Targeted Therapy 7, no. 1: 136. 10.1038/s41392-022-00983-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Vossel, S. , Geng J. J., and Fink G. R.. 2014. “Dorsal and Ventral Attention Systems: Distinct Neural Circuits but Collaborative Roles.” Neuroscientist 20, no. 2: 150–159. 10.1177/1073858413494269. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Yeo, B. T. , Krienen F. M., Sepulcre J., et al. 2011. “The Organization of the Human Cerebral Cortex Estimated by Intrinsic Functional Connectivity.” Journal of Neurophysiology 106, no. 3: 1125–1165. 10.1152/jn.00338.2011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Zhang, H. , Gertel V. H., Cosgrove A. L., and Diaz M. T.. 2021. “Age‐Related Differences in Resting‐State and Task‐Based Network Characteristics and Cognition: A Lifespan Sample.” Neurobiology of Aging 101: 262–272. 10.1016/j.neurobiolaging.2020.10.025. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Data S1: Supporting information.

HBM-47-e70447-s001.docx (112.6KB, docx)

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Articles from Human Brain Mapping are provided here courtesy of Wiley

RESOURCES