Table 3.
Neuroimaging-based brain age studies for general populations or those without neuropsychiatric diagnoses.
| First Author [ref.] |
Year | Cohort | Imaging Modality | ML Algorithm | Main Findings |
|---|---|---|---|---|---|
| Franke [89] | 2013 | 185 people | T1WI | RVR | BAG: +4.6 yr in T2DM, Acceleration by +0.2 yr per year |
| Franke [90] | 2014 | 228 elderly | T1WI | RVR | BAG associated with health markers with gender-specific pattern |
| Franke [91] | 2015 | 8 women | T1WI | RVR | BAG changes during the course of the menstrual cycle |
| Luders [92] | 2016 | 50 LTM, 50 HCs | T1WI | RVR | BAG: −7.5 yr in LTM |
| Cole [21] | 2018 | 669 people | T1WI | GPR | Higher BAG was associated with weaker grip strength, poorer lung function, slower walking speed, lower fluid intelligence, higher allostatic load, and increased mortality risk. |
| Franke [93] | 2018 | 118 elderly | T1WI | RVR | BAG: +4.3 yr in males whose mothers were exposed to famine in early gestation |
| Hatton [94] | 2018 | 359 men | T1WI | SVR | BAG associated with negative fateful life events in midlife |
| Kiehl [95] | 2018 | 1332 incarcerated males | T1WI | ICA | Brain age predicts recidivism, particularly when combined with other data. |
| Le [96] | 2018 | 20 healthy people | T1WI | SVR | BAG: −1.15 or −1.18 yr by taking ibuprofen |
| Luders [97] | 2018 | 14 healthy women after childbirth | T1WI | RVR | Brain age became younger in late postpartum by 5.4 yr. |
| Rogenmoser [98] | 2018 | 42 pro-musician, 45 amateurs, 38HCs | T1WI | RVR | BAG: −3.70 to −4.51 yr in musicians |
| Scheller [99] | 2018 | 34 elderly | T1WI | RVR | interaction of BAG and APOE variants, suggesting a compensation mechanism in the elderly |
| de Lange [100] | 2019 | 12,021 women | T1WI | XGBoost | BAG decrease with the number of previous childbirths |
| Cruz-Almeida [101] | 2019 | 47 elderly | T1WI | GPR | Increased BAG in elderly with chronic pain |
| Cole [15] | 2020 | 14,701 people | T1WI, FLAIR, T2*, DTI, fMRI | LASSO | BAS associated with stroke history, diabetes, smoking, alcohol, and cognitive measures |
| de Lange [102] | 2020 | 473 people | T1WI, DTI, fMRI | XGBoost | Associated with cardiovascular risk |
| de Lange [103] | 2020 | 19,787 women | T1WI | XGBoost | BAG decrease with the number of previous childbirths. Involvement of brain subcortical regions |
| Henneghan [104] | 2020 | 43 breast cancer with chemotherapy, 50 HCs | T1WI | SVR/RF | Trend-level increase on BAG after chemotherapy for breast cancer |
| Reuben [105] | 2020 | 564 people at 45 yr | T1WI | SVR/RF | BAG: +0.77 yr in those who had lead exposure in childhood |
| Seidel [106] | 2020 | 20 sepsis survivors with cognitive deficits, 44 HCs | T1WI | Kernel regression | BAG: +4.5 yr in sepsis survivor, associated with the severity of dyscognition |
| Anaturk [107] | 2021 | 537 elderly | T1WI, DTI, FLAIR | XGBoost | Relationship with cumulative lifestyle measures independent of cognitive age |
| Bittner [108] | 2021 | 622 elderly | T1WI | RVR | BAG: +5.04 months by combined lifestyle risk, +0.6 month by smoking, −0.55 month by physical activity |
| Cherbuin [109] | 2021 | 335 middle age, 351 elderly | T1WI | RVR | BAG: +51.1–65.7days by every additional 10-mmHg increase in BP |
| Dunas [110] | 2021 | 351 people | T1WI, DTI, fMRI | OLS, BRR, LASSO, ENET, SVR, RVR, GPR | BAG associated with current and past physical fitness and cognitive ability |
| Elliott [111] | 2021 | 869 middle-age | T1WI | SVR/RF | Associated with cognitive function, impaired brain health at age 3, and other signs of aging |
| Hedderich [112] | 2021 | 101 premature-born adults, 111 full-term controls | T1WI | RVR | BAG: +1.4 yr in premature-born adults, associated with low gestational age, low birth weight, and increased neonatal treatment intensity |
| Karim [113] | 2021 | 78 older adults | T1WI, T2WI, FLAIR | GPR | BAG associated with male sex, worry, and rumination |
| Rakesh [114] | 2021 | 166 adolescents | T1WI | SVR | increased BAG by neighborhood disadvantage, modulated by effortful control |
| Rosemann [115] | 2021 | 169 elderly | T1WI | GPR | No association with age-related hearing loss |
| Salih [116] | 2021 | 15,335 HCs | DTI | Bayesian ridge regression | Limbic tract-based BAG was most accurate and associated with daily life factors. Two SNPs were associated with BAG. |
| Sanders [117] | 2021 | 122 elderly | T1WI | XGBoost | BAG decrease in more physically active women but not men |
| Subramaniapillai [118] | 2021 | 1067 elderly | T1WI | Elastic net regression | Brain age was more associated with AD risk factors in women than in men. |
| Vidal-Pineiro [119] | 2021 | 6950 people | T1WI | LASSO, XGBoost | No association between cross-sectional brain age and longitudinal change. Association with congenital factors, suggesting a lifelong influence on brain structure from early life |
| Weihs [120] | 2021 | 690 people | T1WI | OLS | Brain age associated with AHI and ODI in PSG data |
| Angebrandt [121] | 2022 | 240 HCs, 231 HCs (middle age) | T1WI | SVR/RF | Dose-dependent relation between 90-day alcohol consumption and BAG |
| Beck [122] | 2022 | 790 healthy people | T1WI, DTI | XGBoost | T1-based BAG: associated with sBP, smoking, pulse, and CRP. DTI-based BAG: associated with phosphate, MCV |
| Bourassa [123] | 2022 | 910 people (midlife) | T1WI | SVR/RF | BAG in midlife is associated with smoking, obesity, and psychological problems during adolescence. |
| Giannakopoulos [124] | 2022 | 80 elderly | T1WI | RVR | BAG predicted a decrease in executive function over time. |
| Linli [125] | 2022 | 33,293 people | T1WI | XGBoost | BAG: +1.19 yr in active regular smokers, associated with the amount of smoking |
| Sone [16] | 2022 | 773 elderly | T1WI | SVR | BAG: associated with life satisfaction, alcohol use, and diabetes |
| Vaughan [126] | 2022 | 57 elderly | T1WI | GPR | BAG: associated with leg strength, moderating the relationship between strength and mobility |
| Wang [127] | 2022 | 165 elderly | T1WI | RVR | BAG: associated with female gender, higher education but not with APOE-e4 or family history of dementia |
| Whistel [128] | 2022 | 712 people | T1WI | SVR | Association of BAG in mid- to late-life with heavier smoking and alcohol consumption in early mid-life |
| Zheng [129] | 2022 | 1676 HCs | T1WI | RBF-kernel | BAG associated with worse cognitive outcomes over time |
BAG: brain age gap, CNN: convolutional neural network, DTI: diffusion tensor imaging, ENET: efficient neural network, fMRI: functional MRI, GPR: Gaussian process regression, HCs: healthy controls, ICA: independent component analysis, LTM: long-term meditation practitioner, ML: machine learning, OLS: ordinary least squares, RBF: radial basis function, RF: random forest, RVR: relevance vector regression, SVR: support vector regression, T1WI: T1-weighted image, T2DM: type 2 diabetes mellitus, T2WI: T2-weighted image.