Esten Høyland Leonardsen

• Roe, James Michael; J., Jagust, William; M., Landau, Susan; M., Harrison, Theresa; Grydeland, Håkon & Slivka, Maksim [Show all 19 contributors for this article] (2026). Cortical thickness changes precede high levels of amyloid by at least 7 years | Nature Neuroscience. Nature Neuroscience. ISSN 1097-6256. doi: 10.1038/s41593-026-02363-4. Full text in Research Archive Show summary

  Alzheimer’s disease is now defined by underlying pathology, with elevated amyloid-beta (Aβ) sufficient for diagnosis in the absence of cognitive symptoms. We combined longitudinal magnetic resonance imaging and Aβ positron emission tomography data from three cognitively healthy cohorts to examine cortical thickness trajectories in individuals who later converted to Aβ-positive status, using magnetic resonance images acquired years before conversion. Individuals who subsequently developed elevated Aβ showed a thicker cortex and reduced cortical thinning, detectable up to 7 years before conversion. Many effects persisted after accounting for quantitative Aβ levels, suggesting some cortical thickness changes may be partly independent of Aβ. Differences in cortical thickness and its change showed moderate spatial correspondence with Aβ deposition patterns, and the timing of thickness changes tracked the progression of Aβ accumulation. These findings indicate that cortical thickness alterations can precede positron emission tomography-detectable amyloid positivity by several years, suggesting high amyloid burden may not represent the earliest imaging marker of Alzheimer’s disease. Longitudinal MRI reveals that the cortex becomes relatively thicker years before amyloid positivity is detected on PET, providing insight into the earliest brain changes and potential consequences of amyloid accumulation in Alzheimer’s disease.

• Helland, Gisle Berg; Ihle-Hansen, Håkon; Aamodt, Anne Hege; Leonardsen, Esten Høyland; Kaufmann, Tobias Herbert & Enriquez, Brian Anthony [Show all 13 contributors for this article] (2026). Brain age trajectories and cognition after stroke in two longitudinal cohorts. Brain Communications. 8(1). doi: 10.1093/braincomms/fcaf509. Full text in Research Archive Show summary

  Brain age is a promising neuroimaging biomarker, reflecting biological aging, but long-term trajectories and predictive value for cognitive outcomes post-stroke remains unclear. This study aimed to characterize brain aging trajectories over 8 years following a first-ever stroke and to evaluate the predictive value of brain age estimates for long-term cognitive outcomes. We analysed data from working-age (<65 years) ischaemic stroke patients with small- and medium-sized strokes (lesion volumes <70 ml), using two longitudinal stroke cohorts. T1-weighted MRI was acquired in the acute phase and at multiple time points up to 8 years post-stroke. Montreal cognitive assessment (MoCA) was assessed at follow-up sessions. Brain age was estimated using a state-of-the-art deep learning model. Brain-predicted age difference (Brain-PAD) was calculated as estimated brain age minus chronological age and corrected by regressing on age, age² and sex. Linear mixed-effects models examined Brain-PAD over multiple time points (whole-brain, ipsilesional and contralesional). Normalized brain volume was derived from FreeSurfer and included in the whole-brain analysis. Linear regression models evaluated whether brain age was associated with cognitive performance (MoCA) at long-term follow-up. We included 120 patients [n = 50 (42%) female, mean ± SD age at discharge was 54.9 ± 9 and National Institutes of Health Stroke Scale was 3.7 ± 6.4], with a mean follow-up of 3.4 ± 2.5 years. The mean MoCA score at follow-up was 24.7 ± 3.7. Brain-PAD increased significantly over time in the whole-brain analysis (β = 0.6/year, P < 0.01), indicating 60% acceleration in brain aging after stroke, with the association remaining significant after adjusting for normalized brain volume (β = 0.5/year, P < 0.01). Accelerated brain aging was observed in the ipsilesional hemisphere (β = 0.7/year, P < 0.01), but not the contralesional hemisphere (β = 0.3/year, P = 0.12). Higher brain age in the acute phase of stroke predicted lower MoCA scores at follow-up (β = −0.12, P < 0.05), whereas chronological age was not a significant predictor (P = 0.12). The association between brain age estimations and cognitive performance remained significant after adjusting for age, sex and education (β = −0.42, P < 0.01). In this longitudinal study, we found accelerated brain aging following stroke. Furthermore, brain age was associated with cognitive outcomes several years later, highlighting its potential as an early biomarker for long-term cognitive prognosis.

• Moskal, Victoria; Voldsbekk, Irene; Leonardsen, Esten Høyland; Holthe, Linn Charlotte Figenschou; Kjelkenes, Rikka & Westlye, Lars Tjelta (2025). Linking cognitive performance and age prediction with psychological distress in youth. Acta Psychologica. ISSN 0001-6918. 261, p. 105795–105795. doi: 10.1016/j.actpsy.2025.105795. Full text in Research Archive
• Fischer-Vieler, Thomas; Rokicki, Jaroslav; Kim, Milin; Leonardsen, Esten Høyland; Wolfers, Thomas & Bell, Christina [Show all 14 contributors for this article] (2025). Mapping cerebellar subregional volumes and heterogeneity in schizophrenia spectrum disorders and violence. European Archives of Psychiatry and Clinical Neuroscience. ISSN 0940-1334. doi: 10.1007/s00406-025-02081-1. Full text in Research Archive Show summary

  Cerebellar abnormalities have been linked to schizophrenia and aggressive behaviour. Subregional cerebellar morphology reflects structural organization and function and show great heterogeneity, which may be linked to shared or distinct underlying mechanisms of violence and psychosis but remain unexplored in these marginalized groups. Total and subregional cerebellar volumes were estimated from 3 T MRI scans from persons with schizophrenia spectrum disorders without (SSD-NV; n = 107) or with (SSD-V; n = 36) a history of severe violence, violent offenders without schizophrenia (NPV; n = 20), and 411 healthy controls. Group difference analyses by GLMs were complemented by novel heterogeneity analyses using pre-trained models and norm-charts of lifespan cerebellar volumes to assess individual deviation patterns. We found decreased grey matter volumes in the posterior cerebellar hemispheres and the vermal regions in both SSD-V and SSD-NV, but with a different subregional distribution. Specifically, SSD-V showed differences in the right Crus I (p = .045, Cohen’s d = .62) and Vermis IX (p < .001, Cohen’s d = .81) in the groupwise comparisons. In addition, individual-level extreme negative deviations were detected in the Vermis IX region in 11% of SSD-V and in some subregions of the posterior cerebellum and the vermis in > 10% of NPV. By applying novel analytical tools, we report different patterns of cerebellar subregion volume reductions associated with violence in individuals with or without schizophrenia spectrum disorders. The partly converging results from the group comparisons and the normative modelling analyses demonstrate the usefulness of complemental methodological approaches to disentangle the complex biological associations between violence and psychosis.

• Skattebøl, Lars; Nygaard, Gro Owren; Leonardsen, Esten Høyland; Kaufmann, Tobias Herbert; Moridi, Thomas & Stawiarz, Leszek [Show all 23 contributors for this article] (2025). Brain age in multiple sclerosis: a study with deep learning and traditional machine learning. Brain Communications. 7(3). doi: 10.1093/braincomms/fcaf152. Full text in Research Archive Show summary

  ‘Brain age’ is a numerical estimate of the biological age of the brain and an overall effort to measure neurodegeneration, regardless of disease type. In multiple sclerosis, accelerated brain ageing has been linked to disability accrual. Artificial intelligence has emerged as a promising tool for the assessment and quantification of the impact of neurodegenerative diseases. Despite the existence of numerous AI models, there is a noticeable lack of comparative imaging data for traditional machine learning versus deep learning in conditions such as multiple sclerosis. A retrospective observational study was initiated to analyse clinical and MRI data (4584 MRIs) from various scanners in a large longitudinal cohort (n = 1516) of people with multiple sclerosis collected from two institutions (Karolinska Institute and Oslo University Hospital) using a uniform data post-processing pipeline. We conducted a comparative assessment of brain age using a deep learning simple fully convolutional network and a well-established traditional machine learning model. This study was primarily aimed to validate the deep learning brain age model in multiple sclerosis. The correlation between estimated brain age and chronological age was stronger for the deep learning estimates (r = 0.90, P < 0.001) than the traditional machine learning estimates (r = 0.75, P < 0.001). An increase in brain age was significantly associated with higher expanded disability status scale scores (traditional machine learning: t = 5.3, P < 0.001; deep learning: t = 3.7, P < 0.001) and longer disease duration (traditional machine learning: t = 6.5, P < 0.001; deep learning: t = 5.8, P < 0.001). No significant inter-model difference in clinical correlation or effect measure was found, but significant differences for traditional machine learning-derived brain age estimates were found between several scanners. Our study suggests that the deep learning-derived brain age is significantly associated with clinical disability, performed equally well to the traditional machine learning-derived brain age measures, and may counteract scanner variability.

• Ebrahimi, Omid Vakili; Sandbakken, Ella Marie; Moss, Sigrun Marie; Johnson, Sverre Urnes; Hoffart, Asle & Bauermeister, Sarah [Show all 9 contributors for this article] (2025). Modifiable risk factors of vaccine hesitancy: insights from a mixed methods multiple population study combining machine learning and thematic analysis during the COVID-19 pandemic. BMC Medicine. 23(1). doi: 10.1186/s12916-025-03953-y. Full text in Research Archive Show summary

  Background Vaccine hesitancy, the delay in acceptance or reluctance to vaccinate, ranks among the top threats to global health. Identifying modifiable factors contributing to vaccine hesitancy is crucial for developing targeted interventions to increase vaccination uptake. Methods This mixed-methods multiple population study utilized gradient boosting machines and thematic analysis to identify modifiable predictors of vaccine hesitancy during the COVID-19 pandemic. Predictors of vaccine hesitancy were investigated in 2926 Norwegian adults (Mage = 37.91, 79.69% female), before the predictive utility of these variables was investigated in an independent sample of 734 adults in the UK (Mage = 40.34, 57.08% female). Two independent teams of authors conducted the machine learning and thematic analyses, blind to each other’s analytic procedures and results. Results The machine learning model performed well in discerning vaccine hesitant (n = 248, 8.48% and n = 109, 14.85%, Norway and UK, respectively) from vaccine uptaking individuals (n = 2678, 91.52% and n = 625, 85.15%), achieving an AUC of 0.94 (AUPRC: 0.72; balanced accuracy: 86%; sensitivity = 0.81; specificity = 0.98) in the Norwegian sample, and an AUC of 0.98 (AUPRC: 0.89; balanced accuracy: 89%; sensitivity = 0.83; specificity = 0.97) in the out-of-sample replication in the UK. The mixed methods investigation identified five categories of modifiable risk tied to vaccine hesitancy, including illusion of invulnerability, doubts about vaccine efficacy, mistrust in official entities, minimization of the societal impact of COVID-19, and health-related fears tied to vaccination. The portrayal of rare incidents across alternative media platforms as fear amplifiers, and the mainstream media’s stigmatizing presentation of unvaccinated individuals, were provided as additional motives underlying vaccine reluctance and polarization. The thematic analysis further revealed information overload, fear of needles, previous negative vaccination experiences, fear of not getting healthcare follow-up after vaccination if needed, and vaccine aversion due to underlying (psychiatric) illness (e.g., eating disorders) as motives underlying vaccine hesitance. Conclusions The identified influential predictors were consistent across two European samples, highlighting their generalizability across European populations. These predictors offer insights about modifiable factors that could be adapted by public health campaigns in mitigating misconceptions and fears related to vaccination toward increasing vaccine uptake. Moreover, the results highlight the media’s responsibility, as mediators of the public perception of vaccines, to minimize polarization and provide accurate portrayals of rare vaccine-related incidents, reducing the risk aggravating fear and reactance to vaccination.

• Kim, Milin; Sharma, Nitin; Leonardsen, Esten Høyland; Rutherford, Saige; Selbæk, Geir & Persson, Karin Ester Torun [Show all 19 contributors for this article] (2025). Predicting Mental and Neurological Illnesses Based on Cerebellar Normative Features. Biological Psychiatry: Global Open Science. 5(5), p. 1–11. doi: 10.1016/j.bpsgos.2025.100541. Full text in Research Archive
• Roe, James Michael; Vidal-Piñeiro, Didac; Sørensen, Øystein; Grydeland, Håkon; Leonardsen, Esten Høyland & Iakunchykova, Olena [Show all 35 contributors for this article] (2024). Brain change trajectories in healthy adults correlate with Alzheimer’s related genetic variation and memory decline across life. Nature Communications. 15(1). doi: 10.1038/s41467-024-53548-z. Full text in Research Archive Show summary

  Throughout adulthood and ageing our brains undergo structural loss in an average pattern resembling faster atrophy in Alzheimer’s disease (AD). Using a longitudinal adult lifespan sample (aged 30-89; 2–7 timepoints) and four polygenic scores for AD, we show that change in AD-sensitive brain features correlates with genetic AD-risk and memory decline in healthy adults. We first show genetic risk links with more brain loss than expected for age in early Braak regions, and find this extends beyond APOE genotype. Next, we run machine learning on AD-control data from the Alzheimer’s Disease Neuroimaging Initiative using brain change trajectories conditioned on age, to identify AD-sensitive features and model their change in healthy adults. Genetic AD-risk linked with multivariate change across many AD-sensitive features, and we show most individuals over age ~50 are on an accelerated trajectory of brain loss in AD-sensitive regions. Finally, high genetic risk adults with elevated brain change showed more memory decline through adulthood, compared to high genetic risk adults with less brain change. Our findings suggest quantitative AD risk factors are detectable in healthy individuals, via a shared pattern of ageing- and AD-related neurodegeneration that occurs along a continuum and tracks memory decline through adulthood.

• Beck, Dani; Lange, Ann-Marie Glasø de; Gurholt, Tiril Pedersen; Voldsbekk, Irene; Maximov, Ivan & Subramaniapillai, Sivaniya [Show all 19 contributors for this article] (2024). Dissecting unique and common variance across body and brain health indicators using age prediction. Human Brain Mapping. ISSN 1065-9471. 45(6). doi: 10.1002/hbm.26685. Full text in Research Archive Show summary

  Ageing is a heterogeneous multisystem process involving different rates of decline in physiological integrity across biological systems. The current study dissects the unique and common variance across body and brain health indicators and parses inter-individual heterogeneity in the multisystem ageing process. Using machine-learning regression models on the UK Biobank data set (N = 32,593, age range 44.6–82.3, mean age 64.1 years), we first estimated tissue-specific brain age for white and gray matter based on diffusion and T1-weighted magnetic resonance imaging (MRI) data, respectively. Next, bodily health traits, including cardiometabolic, anthropometric, and body composition measures of adipose and muscle tissue from bioimpedance and body MRI, were combined to predict ‘body age’. The results showed that the body age model demonstrated comparable age prediction accuracy to models trained solely on brain MRI data. The correlation between body age and brain age predictions was 0.62 for the T1 and 0.64 for the diffusion-based model, indicating a degree of unique variance in brain and bodily ageing processes. Bayesian multilevel modelling carried out to quantify the associations between health traits and predicted age discrepancies showed that higher systolic blood pressure and higher muscle-fat infiltration were related to older-appearing body age compared to brain age. Conversely, higher hand-grip strength and muscle volume were related to a younger-appearing body age. Our findings corroborate the common notion of a close connection between somatic and brain health. However, they also suggest that health traits may differentially influence age predictions beyond what is captured by the brain imaging data, potentially contributing to heterogeneous ageing rates across biological systems and individuals.

• Helland, Gisle Berg; Beyer, Mona K.; Enriquez, Brian Anthony; Ihle-Hansen, Hege Beate; Ihle-Hansen, Håkon & Andersson, Stein [Show all 13 contributors for this article] (2024). Lesion Volume as a Predictor for Return to Work After Endovascular Treatment: A 4‐Year Prospective Cohort Study. Stroke: Vascular and Interventional Neurology. 4(6). doi: 10.1161/SVIN.124.001494. Full text in Research Archive Show summary

  Background There are no known objective biomarkers for predicting return to work (RTW) in ischemic stroke survivors. This study aims to explore the predictive utility and defines a cutoff value of lesion volume on RTW after endovascular treatment. Methods We included patients aged <65 years undergoing endovascular treatment at Oslo University Hospital between January 2017 and May 2019. Employment status was obtained at both baseline and a 4‐year follow‐up. Stroke lesion volumes were segmented using magnetic resonance imaging scans 24 hours after endovascular treatment. Logistic regression models were conducted to assess the impact of lesion volume on RTW status at follow‐up, adjusted for patients’ characteristics, stroke‐related factors, and treatment. We calculated the receiver operating characteristic curve to determine the optimal lesion volume cutoff. Machine learning regression models were used to assess the predictive abilities of baseline clinical and imaging variables for RTW. Results Of the 109 individuals treated, 81 (74%) were employed at baseline. Among these, 60 completed 4‐year follow‐up with magnetic resonance imaging available for stroke lesion segmentation and were included in the analyses. Mean age at stroke onset was 51.8 years (range, 23.5–64.9 years), and 50% were female. Median lesion volume was 18 mL (interquartile ranges, 45—47 mL). After 4 years, 34 (57%) had successfully RTW. The odds for not RTW increased by 5% for every 1 mL increase in lesion volume (adjusted odds ratio, 1.05 [95% CI, 1.02–1.11]; P = 0.02). A lesion volume cutoff value of 29 cm3 yielded a sensitivity of 0.91 and specificity of 0.65 for predicting RTW. Notably, the most influential feature in the machine learning model for predicting RTW was lesion volume. Conclusion Lesion volume was the most robust predictor of RTW 4 years after endovascular treatment. Our findings suggest that a cutoff of 29 cm3 is suitable to distinguish between those with high and low chance of RTW.

• Dahl, Andreas; Eilertsen, Espen Moen; Cabello, Sara Fernandez Rodriguez; Norbom, Linn B.; Tandberg, Anneli Desire & Leonardsen, Esten [Show all 11 contributors for this article] (2024). Genetic and brain similarity independently predict childhood anthropometrics and neighborhood socioeconomic conditions. Developmental Cognitive Neuroscience. ISSN 1878-9293. 65. doi: 10.1016/j.dcn.2023.101339. Full text in Research Archive Show summary

  Linking the developing brain with individual differences in clinical and demographic traits is challenging due to the substantial interindividual heterogeneity of brain anatomy and organization. Here we employ an integrative approach that parses individual differences in both cortical thickness and common genetic variants, and assess their effects on a wide set of childhood traits. The approach uses a linear mixed model framework to obtain the unique effects of each type of similarity, as well as their covariance. We employ this approach in a sample of 7760 unrelated children in the ABCD cohort baseline sample (mean age 9.9, 46.8% female). In general, associations between cortical thickness similarity and traits were limited to anthropometrics such as height, weight, and birth weight, as well as a marker of neighborhood socioeconomic conditions. Common genetic variants explained significant proportions of variance across nearly all included outcomes, although estimates were somewhat lower than previous reports. No significant covariance of the effects of genetic and cortical thickness similarity was found. The present findings highlight the connection between anthropometrics as well as neighborhood socioeconomic conditions and the developing brain, which appear to be independent from individual differences in common genetic variants in this population-based sample.

• Engvig, Andreas; Kalleberg, Karl Trygve; Westlye, Lars Tjelta; Leonardsen, Esten Høyland & Initiative, Alzheimer's Disease Neuroimaging (2024). Complementary value of molecular, phenotypic, and functional aging biomarkers in dementia prediction. GeroScience. ISSN 2509-2715. p. 1–20. doi: 10.1007/s11357-024-01376-w. Full text in Research Archive Show summary

  DNA methylation age (MA), brain age (BA), and frailty index (FI) are putative aging biomarkers linked to dementia risk. We investigated their relationship and combined potential for prediction of cognitive impairment and future dementia risk using the ADNI database. Of several MA algorithms, DunedinPACE and GrimAge2, associated with memory, were combined in a composite MA alongside BA and a data-driven FI in predictive analyses. Pairwise correlations between age- and sex-adjusted measures for MA (aMA), aBA, and aFI were low. FI outperformed BA and MA in all diagnostic tasks. A model including age, sex, and aFI achieved an area under the curve (AUC) of 0.94 for differentiating cognitively normal controls (CN) from dementia patients in a held-out test set. When combined with clinical biomarkers (apolipoprotein E ε4 allele count, memory, executive function), a model including aBA and aFI predicted 5-year dementia risk among MCI patients with an out-of-sample AUC of 0.88. In the prognostic model, BA and FI offered complementary value (both βs 0.50). The tested MAs did not improve predictions. Results were consistent across FI algorithms, with data-driven health deficit selection yielding the best performance. FI had a stronger adverse effect on prognosis in males, while BA’s impact was greater in females. Our findings highlight the complementary value of BA and FI in dementia prediction. The results support a multidimensional view of dementia, including an intertwined relationship between the biomarkers, sex, and prognosis. The tested MA’s limited contribution suggests caution in their use for individual risk assessment of dementia.

• Kim, Milin; Leonardsen, Esten Høyland; Rutherford, Saige; Selbæk, Geir; Persson, Karin Ester Torun & Steen, Nils Eiel [Show all 15 contributors for this article] (2024). Mapping cerebellar anatomical heterogeneity in mental and neurological illnesses. Nature Mental Health. 2(10), p. 1196–1207. doi: 10.1038/s44220-024-00297-z. Full text in Research Archive Show summary

  The cerebellum is linked to motor coordination, cognitive and affective processing, in addition to a wide range of clinical illnesses. To enable robust quantification of individual cerebellar anatomy relative to population norms, we mapped the normative development and aging of the cerebellum across the lifespan using brain scans of >54,000 participants. We estimated normative models at voxel-wise spatial precision, enabling integration with cerebellar atlases. Applying the normative models in independent samples revealed substantial heterogeneity within five clinical illnesses: autism spectrum disorder, mild cognitive impairment, Alzheimer disease, bipolar disorder, and schizophrenia. Notably, individuals with autism spectrum disorder and mild cognitive impairment exhibited increased positive and negative extreme deviations in cerebellar anatomy, while those with schizophrenia and Alzheimer disease showed predominantly negative deviations. Finally, extreme deviations were associated with cognitive scores. Ou

• Grødem, Edvard Olaf Sætersdal; Leonardsen, Esten Høyland; MacIntosh, Bradley; Bjørnerud, Atle; SCHELLHORN, TILL & Sørensen, Øystein [Show all 8 contributors for this article] (2024). A minimalistic approach to classifying Alzheimer's disease using simple and extremely small convolutional neural networks. Journal of Neuroscience Methods. ISSN 0165-0270. 411. doi: 10.1016/j.jneumeth.2024.110253. Full text in Research Archive
• Iakunchykova, Olena; Leonardsen, Esten Høyland & Wang, Yunpeng (2024). Genetic evidence for causal effects of immune dysfunction in psychiatric disorders: where are we? Translational Psychiatry. 14(1), p. 1–11. doi: 10.1038/s41398-024-02778-2. Full text in Research Archive
• Lange, Ann-Marie Glasø de; Leonardsen, Esten Høyland; Barth, Claudia; Schindler, Louise S; Crestol, Arielle & Holm, Madelene Christin [Show all 10 contributors for this article] (2024). Parental status and markers of brain and cellular age: A 3D convolutional network and classification study. Psychoneuroendocrinology. ISSN 0306-4530. 165. doi: 10.1016/j.psyneuen.2024.107040. Full text in Research Archive
• Leonardsen, Esten Høyland; Persson, Karin Ester Torun; Grødem, Edvard Olaf Sætersdal; Dinsdale, Nicola; SCHELLHORN, TILL & Roe, James Michael [Show all 16 contributors for this article] (2024). Constructing personalized characterizations of structural brain aberrations in patients with dementia using explainable artificial intelligence. npj Digital Medicine. 7(1). doi: 10.1038/s41746-024-01123-7. Full text in Research Archive Show summary

  Deep learning approaches for clinical predictions based on magnetic resonance imaging data have shown great promise as a translational technology for diagnosis and prognosis in neurological disorders, but its clinical impact has been limited. This is partially attributed to the opaqueness of deep learning models, causing insufficient understanding of what underlies their decisions. To overcome this, we trained convolutional neural networks on structural brain scans to differentiate dementia patients from healthy controls, and applied layerwise relevance propagation to procure individual-level explanations of the model predictions. Through extensive validations we demonstrate that deviations recognized by the model corroborate existing knowledge of structural brain aberrations in dementia. By employing the explainable dementia classifier in a longitudinal dataset of patients with mild cognitive impairment, we show that the spatially rich explanations complement the model prediction when forecasting transition to dementia and help characterize the biological manifestation of disease in the individual brain. Overall, our work exemplifies the clinical potential of explainable artificial intelligence in precision medicine.

• Korbmacher, Max; Wang, Mengyun; Eikeland, Rune; Buchert, Ralph; Andreassen, Ole & Espeseth, Thomas [Show all 10 contributors for this article] (2023). Considerations on brain age predictions from repeatedly sampled data across time. Brain and Behavior. doi: 10.1002/brb3.3219. Full text in Research Archive Show summary

  Introduction Brain age, the estimation of a person's age from magnetic resonance imaging (MRI) parameters, has been used as a general indicator of health. The marker requires however further validation for application in clinical contexts. Here, we show how brain age predictions perform for the same individual at various time points and validate our findings with age-matched healthy controls. Methods We used densely sampled T1-weighted MRI data from four individuals (from two densely sampled datasets) to observe how brain age corresponds to age and is influenced by acquisition and quality parameters. For validation, we used two cross-sectional datasets. Brain age was predicted by a pretrained deep learning model. Results We found small within-subject correlations between age and brain age. We also found evidence for the influence of field strength on brain age which replicated in the cross-sectional validation data and inconclusive effects of scan quality. Conclusion The absence of maturation effects for the age range in the presented sample, brain age model bias (including training age distribution and field strength), and model error are potential reasons for small relationships between age and brain age in densely sampled longitudinal data. Clinical applications of brain age models should consider of the possibility of apparent biases caused by variation in the data acquisition process.

• Leonardsen, Esten Høyland; Vidal-Piñeiro, Didac; Roe, James Michael; Frei, Oleksandr; Shadrin, Alexey & Iakunchykova, Olena [Show all 14 contributors for this article] (2023). Genetic architecture of brain age and its casual relations with brain and mental disorders. Molecular Psychiatry. ISSN 1359-4184. doi: 10.1038/s41380-023-02087-y. Full text in Research Archive Show summary

  Abstract The difference between chronological age and the apparent age of the brain estimated from brain imaging data—the brain age gap (BAG)—is widely considered a general indicator of brain health. Converging evidence supports that BAG is sensitive to an array of genetic and nongenetic traits and diseases, yet few studies have examined the genetic architecture and its corresponding causal relationships with common brain disorders. Here, we estimate BAG using state-of-the-art neural networks trained on brain scans from 53,542 individuals (age range 3–95 years). A genome-wide association analysis across 28,104 individuals (40–84 years) from the UK Biobank revealed eight independent genomic regions significantly associated with BAG ( p  &lt; 5 × 10 −8 ) implicating neurological, metabolic, and immunological pathways – among which seven are novel. No significant genetic correlations or causal relationships with BAG were found for Parkinson’s disease, major depressive disorder, or schizophrenia, but two-sample Mendelian randomization indicated a causal influence of AD ( p  = 7.9 × 10 −4 ) and bipolar disorder ( p  = 1.35 × 10 −2 ) on BAG. These results emphasize the polygenic architecture of brain age and provide insights into the causal relationship between selected neurological and neuropsychiatric disorders and BAG.

• Persson, Karin Ester Torun; Leonardsen, Esten Høyland; Edwin, Trine Holt; Knapskog, Anne-Brita; Tangen, Gro Gujord & Selbæk, Geir [Show all 9 contributors for this article] (2023). Diagnostic accuracy of brain age prediction in a memory clinic population and comparison with clinically available volumetric measures. Scientific Reports. 13. doi: 10.1038/s41598-023-42354-0. Full text in Research Archive Show summary

  The aim of this study was to assess the diagnostic validity of a deep learning-based method estimating brain age based on magnetic resonance imaging (MRI) and to compare it with volumetrics obtained using NeuroQuant (NQ) in a clinical cohort. Brain age prediction was performed on minimally processed MRI data using deep convolutional neural networks and an independent training set. The brain age gap (difference between chronological and biological age) was calculated, and volumetrics were performed in 110 patients with dementia (Alzheimer’s disease, frontotemporal dementia (FTD), and dementia with Lewy bodies), and 122 with non-dementia (subjective and mild cognitive impairment). Area-under-the-curve (AUC) based on receiver operating characteristics and logistic regression analyses were performed. The mean age was 67.1 (9.5) years and 48.7% (113) were females. The dementia versus non-dementia sensitivity and specificity of the volumetric measures exceeded 80% and yielded higher AUCs compared to BAG. The explained variance of the prediction of diagnostic stage increased when BAG was added to the volumetrics. Further, BAG separated patients with FTD from other dementia etiologies with > 80% sensitivity and specificity. NQ volumetrics outperformed BAG in terms of diagnostic discriminatory power but the two methods provided complementary information, and BAG discriminated FTD from other dementia etiologies.

• Holm, Madelene Christin; Leonardsen, Esten Høyland; Beck, Dani; Dahl, Andreas; Kjelkenes, Rikka & Lange, Ann-Marie Glasø de [Show all 7 contributors for this article] (2023). Linking brain maturation and puberty during early adolescence using longitudinal brain age prediction in the ABCD cohort. Developmental Cognitive Neuroscience. ISSN 1878-9293. 60. doi: 10.1016/j.dcn.2023.101220. Full text in Research Archive Show summary

  The temporal characteristics of adolescent neurodevelopment are shaped by a complex interplay of genetic, biological, and environmental factors. Using a large longitudinal dataset of children aged 9–13 from the Adolescent Brain Cognitive Development (ABCD) study we tested the associations between pubertal status and brain maturation. Brain maturation was assessed using brain age prediction based on convolutional neural networks and minimally processed T1-weighted structural MRI data. Brain age prediction provided highly accurate and reliable estimates of individual age, with an overall mean absolute error of 0.7 and 1.4 years at the two timepoints respectively, and an intraclass correlation of 0.65. Linear mixed effects (LME) models accounting for age and sex showed that on average, a one unit increase in pubertal maturational level was associated with a 2.22 months higher brain age across time points (β = 0.10, p < .001). Moreover, annualized change in pubertal development was weakly related to the rate of change in brain age (β = .047, p = 0.04). These results demonstrate a link between sexual development and brain maturation in early adolescence, and provides a basis for further investigations of the complex sociobiological impacts of puberty on life outcomes.

• Leonardsen, Esten Høyland; Peng, Han; Kaufmann, Tobias; Agartz, Ingrid; Andreassen, Ole & Celius, Elisabeth Gulowsen [Show all 19 contributors for this article] (2022). Deep neural networks learn general and clinically relevant representations of the ageing brain. NeuroImage. ISSN 1053-8119. 256. doi: 10.1016/j.neuroimage.2022.119210. Full text in Research Archive Show summary

  The discrepancy between chronological age and the apparent age of the brain based on neuroimaging data — the brain age delta — has emerged as a reliable marker of brain health. With an increasing wealth of data, approaches to tackle heterogeneity in data acquisition are vital. To this end, we compiled raw structural magnetic resonance images into one of the largest and most diverse datasets assembled (n=53542), and trained convolutional neural networks (CNNs) to predict age. We achieved state-of-the-art performance on unseen data from unknown scanners (n=2553), and showed that higher brain age delta is associated with diabetes, alcohol intake and smoking. Using transfer learning, the intermediate representations learned by our model complemented and partly outperformed brain age delta in predicting common brain disorders. Our work shows we can achieve generalizable and biologically plausible brain age predictions using CNNs trained on heterogeneous datasets, and transfer them to clinical use cases.

• Vidal-Piñeiro, Didac; Wang, Yunpeng; Krogsrud, Stine Kleppe; Amlien, Inge; Baaré, William FC & Bartres-Faz, David [Show all 31 contributors for this article] (2021). Individual variations in ‘brain age’ relate to early-life factors more than to longitudinal brain change. eLIFE. 10. doi: 10.7554/eLife.69995. Full text in Research Archive Show summary

  Brain age is a widely used index for quantifying individuals’ brain health as deviation from a normative brain aging trajectory. Higher-than-expected brain age is thought partially to reflect above-average rate of brain aging. Here, we explicitly tested this assumption in two independent large test datasets (UK Biobank [main] and Lifebrain [replication]; longitudinal observations ≈ 2750 and 4200) by assessing the relationship between cross-sectional and longitudinal estimates of brain age. Brain age models were estimated in two different training datasets (n ≈ 38,000 [main] and 1800 individuals [replication]) based on brain structural features. The results showed no association between cross-sectional brain age and the rate of brain change measured longitudinally. Rather, brain age in adulthood was associated with the congenital factors of birth weight and polygenic scores of brain age, assumed to reflect a constant, lifelong influence on brain structure from early life. The results call for nuanced interpretations of cross-sectional indices of the aging brain and question their validity as markers of ongoing within-person changes of the aging brain. Longitudinal imaging data should be preferred whenever the goal is to understand individual change trajectories of brain and cognition in aging.

View all works in NVA

• Abdulkadir, Ahmed; Dvornek, Nicha C.; Habes, Mohamad; Leonardsen, Esten Høyland; Xiao, Yiming & Bathula, Deepti R. [Show all 9 contributors for this article] (2023). Preface. Lecture Notes in Computer Science (LNCS). ISSN 0302-9743. 14312, p. v–vi. Full text in Research Archive
• Korbmacher, Max; Wang, Meng-yun; Eikeland, Rune; Buchert, Ralph; Andreassen, Ole & Espeseth, Thomas [Show all 10 contributors for this article] (2023). Field strength and potentially scan quality influence brain age estimates. Full text in Research Archive
• Roe, James Michael & Leonardsen, Esten Høyland (2023). Brain age as a marker of overall brain health: Phenotypic associations, genetic foundations and causal relationships. Full text in Research Archive

View all works in NVA