Table 1.
Studies of brain-predicted age in disease
| Study | Clinical group | N | Age (mean ± SD, or range) | MRI features | Algorithm | Brain-PAD (mean, years) |
|---|---|---|---|---|---|---|
| Psychiatric disorders | ||||||
| Franke et al. [21] | Alzheimer’s disease | 102 | 76 ± 8 | GM | RVR | 10.0 |
| Franke and Gaser [46] | MCI—stable | 36 | 77 ± 6 | GM | RVR | BL: −0.5 FU (3 yrs): −0.4 |
| MCI—progressive | 112 | 74 ± 7 | GM | RVR | BL: 6.2 FU (3 yrs): 9.0 | |
| Alzheimer’s disease | 150 | 75 ± 8 | GM | BL: 6.7 FU (2 yrs): 9.0 | ||
| Koutsouleris et al. [24] | High psychosis risk | 89 | 25 ± 6 | GM | SVR | 1.7 |
| Schizophrenia | 141 | 28 ± 12 | GM | SVR | 5.5 | |
| Major depression | 104 | 42 ± 8 | GM | SVR | 4.0 | |
| Gaser et al. [49] | MCI—progressive (early/late) | 58/75 | 74 ± 7/75 ± 7 | GM | RVR | 8.7/5.6 |
| Schnack et al. [29] | Schizophrenia | 341 | 34 ± 12 | GM | SVR | BL: 3.4 FU (4 yrs): 4.3 |
| Löwe et al. [48] | MCI—stable (APOE ε4 carriers/non-carriers) | 14/22 | 77 ± 6/77 ± 6 | GM | RVR | BL: −0.9/−0.9 FU(3 yrs): 0.0/−0.6 |
| MCI—progressive (APOE ε4 carriers/non-carriers) | 78/34 | 74 ± 6/75 ± 9 | GM | RVR | BL: 5.8/5.5 FU (3 yrs): 8.7/7.3 | |
| Alzheimer’s disease (APOE ε4 carriers/non-carriers) | 101/49 | 74 ± 7/76 ± 9 | GM | RVR | BL: 5.8/6.2 FU (2 yrs): 8.3/7.7 | |
| Nenadic et al. [44] | Bipolar disorder | 22 | 38 ± 11 | GM | SVR | −1.3 |
| Borderline personality disorder | 57 | 26 ± 7 | GM | SVR | 3.1 | |
| Schizophrenia | 45 | 34 ± 10 | GM | SVR | 2.6 | |
| Li et al. [47] | Alzheimer’s disease | 411 | 75 ± 7 | Hippocampal volume | SVR | 7.0 |
| Varikuti et al. [66] | Alzheimer’s disease | 163 | 56–91 | GM | LASSO | 8.5; 10.7a |
| MCI | 64 | 55–87 | GM | LASSO | 6.2; 5.4a | |
| Kolenic et al. [131] | Psychosis (first episode) | 120 | 27 ± 4.9 | GM | RVR | 2.6 |
| Guggenmos et al. [132] | Alcohol dependence | 119 | 20–65 | GM | MLRR | 4.0 |
| Neurological disorders | ||||||
| Cole et al. [52] | Traumatic brain injury | 99 | 38 ± 12 | GM/WM | GPR | 4.7/6.0 |
| Cole et al. [59] | HIV | 162 | 57 ± 8 | Whole brain | GPR | 2.2 |
| Cole et al. [133] | HIV | 131 | 56 ± 6 | Whole brain | GPR | BL: 1.6 FU (2 yrs): 1.6 |
| Cole et al. [50] | Down’s syndrome | 46 | 42 ± 9 | Whole brain | GPR | 2.5 |
| Pardoe et al. [58] | Epilepsy (medically refractory/newly-diagnosed) | 94/42 | 32 ± 14/31 ± 11 | Whole brain | GPR | 4.5/0.9 |
| Liem et al. [53] | Objective cognitive impairment (mild/major) | 632/251 | 58 ± 15/58 ± 16 | Whole brain | SVR/RF | 0.7/1.7 |
| Physiological disorders | ||||||
| Franke et al. [60] | Diabetes (type II) | 98 | 65 ± 8 | GM | RVR | 4.6 |
| Diabetes (type II)—longitudinal | 12 | 63 ± 7 | GM | RVR | BL: 5.1 FU (4 yrs): 5.9 | |
| Ronan et al. [61] | Obesity | 227 | 58 ± 17 | WM | NLME | 10.0 |
| Franke et al. [134] | Gestational nutrient restriction (female/male) | 22/19 | 67 ± 0.2/67 ± 0.1 | GM | RVR | 0.9/2.5 |
BL baseline, FU follow-up, GM grey matter, GPR Gaussian process regression, LASSO Least Absolute Shrinkage and Selection Operator, MCI mild cognitive impairment, MLRR multi-linear ridge regression, NLME non-linear mixed effects model, RF random forests, RVR relevance vector regression, SVR support vector regression, WM white matter
aStudy included results from two different training datasets