TABLE 1.
Summary of studies.
| Study | Diagnosis | Patient N | Control N | BDNF type | BDNF in patients versus controls | BDNF ES (95% CI) | Cognition measures | Cognition in patients versus controls | BDNF-cognition relationship | Level of evidence (54) |
| Asevedo et al. (59) | SZ | 30 | 27 | Plasma | ↑ | 1.04 (0.52; 1.57) |
Verbal learning, verbal fluency, working memory, set shifting, inhibition, executive function tests | Deficits in verbal learning | BDNF levels positively correlated with semantic generation tasks | B |
|
| ||||||||||
| Carlino et al. (42) | CH-SZ | 40 | 40 | Serum | ↓ | – | Processing speed, attention, executive function, working memory tests | Significantly poorer neurocognitive performance | Serum truncated-BDNF abundance predicted for high cognitive deficits | B |
|
| ||||||||||
| Chang et al. (55) | BD-II | 228 | 135 | Plasma | ↓ | −0.23 (−0.44; −0.01) |
WMS | Significantly lower scores on 5 subtests of WMS | BDNF more likely to be associated with clinical characteristics than with memory | B |
|
| ||||||||||
| Chou et al. (49) |
EU-BD-I | 23 | 33 | Plasma | NS | −0.02 (−0.55; 0.51) |
Attention, memory, executive function tests | Cognitive deficits present | Deficits in cognition not significantly correlated with BDNF except two items from tests | B |
|
| ||||||||||
| Dell’Osso et al. (53) | D-BD-I | 16 | 15 | Plasma | ↓ | −1.71 (−2.55; −0.87) |
Cognitive disturbances factor score form HRSD | NA | BDNF levels may be related to severity of depression and retardation symptoms | B |
|
| ||||||||||
| Dias et al. (52) | EU-BD-I | 65 | 50 | Serum | NS | 0.19 (−0.18; 0.56) |
Attention and mental control, perceptual-motor skills, executive functions, verbal fluency and abstraction, visuospatial attention, memory tests | Significantly worse results on 11 out of the 16 neurocognitive tests | Significant positive association between serum BDNF levels and a test of verbal fluency in both BD patients and controls | B |
|
| ||||||||||
| Dong et al. (64) | SZ | 818 | 467 | Serum | ↓ | CLZ, MA: −0.85 (−1.02; −0.67) CLZ, FE: −0.13 (−0.40; 0.15) RISP, MA: −0.86 (−1.08; −0.64) RISP, FE: −0.55 (−0.86; −0.24) TYP, MA: −0.99 (−1.19; −0.78) TYP, FE: −0.09 (−0.47; 0−29) |
RBANS | Significantly lower scores | Association between BDNF and cognitive performance in only male patients and female patients taking typical antipsychotics | B |
|
| ||||||||||
| Hori et al. (46) | CH-SZ | 86 | 51 | Serum | NS | −0.32 (−0.67; 0.03) | IGT | Significantly lower scores in IGT | Negative correlation between BDNF levels and mean net scores on the trials in the final two blocks | B |
|
| ||||||||||
| Hori et al. (60) | SZ | 146 | 51 | Serum | NS | −0.24 (−0.56; 0.08) | BACS | NA | Negative correlations between serum BDNF levels and scores for verbal memory, attention and processing speed | B |
|
| ||||||||||
| Li et al. (70) | SZ | 472 | 225 | Serum | ↓ | −1.56 (−1,72; −1.40) | RBANS | Significantly lower RBANS total score | Serum BDNF independently positively correlated with attention and immediate memory | B |
|
| ||||||||||
| Man et al. (34) | FEP-SZ | 80 | 80 | Serum | ↓ | −1.22 (−1.56; −0.88) | RBANS | Significantly lower cognitive performance on the RBANS total and four of its five subscale scores | No significant correlation between BDNF and any index or total scores of RBANS | B |
|
| ||||||||||
| Mora et al. (50) | EU-BD and MA-BD | 84 (52 EU; 32 MA) | 49 | Serum | ↓ | EU: −0.50 (−0.89; −0.10) MA: −0.89 (−1.35; −0.44) |
Executive function, selective attention, inhibition, processing speed, cognitive flexibility, sustained attention, perseverative behavior, verbal learning, recall, recognition, visual memory tests | Worse performance in executive functioning, inhibition, processing speed, verbal and visual memory | BDNF levels associated with executive functioning and verbal memory, together with other demographic variables | B |
|
| ||||||||||
| Niitsu et al. (61) | SZ | 63 | 52 | Serum | NS | 0.17 (0,20; 0.54) |
WAIS-R, VFT, WCST, TMT, Stroop test, DSDT | Significantly worse performance on all tests | Serum BDNF levels related to the impairment of verbal working memory in patients | B |
|
| ||||||||||
| Penadés et al. (62) | SZ | 70 | 15 | Serum | NS | 0.25 (−0.31; 0.81) | Global cognition, working memory, processing speed, verbal memory, non-verbal memory, executive function tests | Significantly worse performance on most tests | No significant correlation between serum BDNF level and cognition | A |
|
| ||||||||||
| Qu et al. (35) | DN-FEP-SZ | 256 | 177 | Serum | ↓ | M: −1.07 (−1.28; −0.86) F: −0.85 (−1.09; −0.62) |
RBANS | Cognitive function decreased | No association between BDNF and cognitive function | B |
|
| ||||||||||
| de Azua et al. (36) | FEP-SZ | 45 | 45 | Plasma | ↓ | −0.78 (−1.20; −0.35) | Learning ability, immediate and delayed memory, abstract thinking, and processing speed tests | Cognitive performance of patients significantly worse | Plasma BDNF levels at 6 months after first hospitalization positively associated with several cognitive domains | B |
|
| ||||||||||
| Rybakowski et al. (51) | EU-BD-I | 60 | 60 | Plasma | ↓ | −0.43 (−0.80; −0.07) | CANTAB | Lithium-treated patients had poorer results on all domains of neuropsychological tests | Performance on neuropsychological tests and plasma BDNF levels in excellent lithium responders is different compared to patients lacking the optimal effect of lithium but not different compared to matched healthy controls | B |
|
| ||||||||||
| Tang et al. (82) | SZ | 109 | 40 | Serum | ↓ | DS: −2.44 (−2.86; −2.02) NDS: −2.25 (−2.66; −1.84) |
Processing speed, attention, executive function, working memory tests | Significantly worse performance | No correlations between BDNF levels and the cognitive tests in SZ and HC groups | B |
|
| ||||||||||
| Theleritis et al. (37) | FEP-SZ | 87 | 152 | Plasma | NS | 0.26 (−0.01; 0−52) | IQ, verbal memory and learning, visual memory, executive function, working memory, attention, concentration, processing speed, verbal fluency tests | NA | No association between BDNF plasma levels and cognitive functions | B |
|
| ||||||||||
| Vinogradov et al. (43) | CH-SZ | 56 | 15 | Serum | ↓ | −0.59 (−1.16; −0.03) | MCCB | Decrement in cognitive functioning (∼ 1 SD below the normal mean) | No significant association between change in BDNF and change in global cognition | A |
|
| ||||||||||
| Wei et al. (44) | CH-SZ | 189 | 60 | Serum | ↓ | −1.00 (−1.30; −0.70) | Executive function test | Executive function impaired | BDNF may be a useful biomarker for executive dysfunction | B |
|
| ||||||||||
| Wu et al. (45) | CH-SZ | 83 | 52 | Serum | ↓ | TD: −1.00 (−1.43; −0.57) WTD: −0.65 (−1.05; −0.25) |
RBANS | Significantly lower scores in almost all subscales | No significant associations between BDNF and RBANS total score or any cognitive index | B |
|
| ||||||||||
| Wu et al. (38) | DN-FEP-SZ | 354 | 152 | Serum | ↓ | −1.08 (−1.27; −0.89) | RBANS | Extensive cognitive impairment | No significant association between BDNF levels and RBANS total score or its index scores | B |
|
| ||||||||||
| Xiao et al. (63) | DN-FEP-SZ | 58 | 55 | Serum | ↓ | −0.99 (−1.39; −0.65) | Verbal fluency, attention and processing speed, attention distribution, working memory, motor speed, and executive function tests | Significantly worse on nearly all neurocognitive tests | BDNF levels positively correlated with the animal subscale of the VFT and negatively correlated with TMT-part B scores | B |
|
| ||||||||||
| Xiu et al. (39) | SZ | 232 | 60 | Serum | ↓ | −0.81 (−1.10; −0.53) | Executive function tests | Significantly lower scores | Lower BDNF levels were correlated with executive dysfunction | B |
|
| ||||||||||
| Xiu et al. (56) | DN-FEP-SZ | 327 | 391 | Serum | ↓ | −0.88 (−1.04; −0.73) | RBANS | Significantly lower scores | No relationship between BDNF and cognitive impairments | B |
|
| ||||||||||
| Yang et al. (40) | FEP-SZ, CH-SZ | 65 34 FEP, 31 CH |
35 | Plasma | ↓ | FEP: −0.44 (−0.91; 0.04) CH: −0.62 (−1.11; −0.13) |
MCCB | Index scores remarkably lower | Low BDNF levels were associated with cognitive impairments | B |
|
| ||||||||||
| Zhang et al. (69) | SZ | 575 | 405 | Serum | ↓ | Val/Val: −0.77 (−1.02; −0.52) Val/Met: −0.82 (−1.00; −0.65) Met/Met: −0.88 (−1.15; −0.60) |
RBANS | Significantly lower in cognitive scores in nearly all subscales | Higher serum BDNF levels were associated with better cognitive function | B |
|
| ||||||||||
| Zhang et al. (48) | CH-SZ | 251 | 206 | Serum | ↓ | −0.93 (−1.13; −0.74) | RBANS | Significantly lower scores | BDNF positively associated with immediate memory | B |
|
| ||||||||||
| Zhang et al. (47) | CH-SZ | 248 | 188 | Serum | ↓ | −0.91 (−1.10; −0.72) | RBANS | Worse performance on most of the cognitive tasks | BDNF positively associated with immediate memory in female patients | B |
|
| ||||||||||
| Zhang et al. (57) | SZ | 108 | 47 | Serum | ↓ | −1.70 (−2,05; −1.36) | RBANS | Significantly lower scores | Metabolic adverse effects of olanzapine may aggravate cognitive dysfunction in patients with schizophrenia through an interaction between BDNF | B |
|
| ||||||||||
| Zhang et al. (58) | AC-SZ | 68 | 47 | Plasma | ↓ | −0.52 (−0.89; −0.14) | RBANS | Decreased compared to controls | Increase in plasma levels of BDNF significantly correlated with the change in the RBANS total scores | B |
AC, acute; BACS, Brief Assessment of Cognition in Schizophrenia; BD, bipolar disorder; BDNF, brain-derived neurotrophic factor; CANTAB, Cambridge Neuropsychological Test Automated Battery; CH, chronic; CLZ, clozapine; D, depressed; DN, drug-naïve; DS, deficit schizophrenia; DSDT, Digit Span Distraction Test; ES, effect size; EU, euthymic; F, female; FEP, first episode; HRSD, Hamilton Rating Scale for Depression; IGT, Iowa Gambling Task; MA, manic; M, male; MCCB, MATRICS Consensus Cognitive Battery; NDS, non-deficit schizophrenia; NA, not available, NS, not significant; RBANS, Repeatable Battery for the Assessment of Neuropsychological; RISP, risperidone; SZ, schizophrenia; TD, tardive dyskinesia; TMT, Trail Making Test; TYP, typical; VFT, Verbal Fluency Test; WAIS-R, Wechsler Adult Intelligence Scale Revised; WCST, Wisconsin Card Sorting Test; WMS, Wechsler Memory Scale; WTD, without tardive dyskinesia.