Decreased levels of brain-derived neurotrophic factor (BDNF) are assumed to play a crucial role in the pathophysiology of mild neurocognitive disorders (MNCDs). In this study, we compared plasma BDNF levels (at baseline and after two months of treatment with escitalopram) in patients with the main types of MNCDs and normal controls. 21 patients met the DSM-5 diagnostic criteria for possible MNCD due to Alzheimer’s disease (MNCD-AD); 22 patients fulfilled the diagnostic criteria for subcortical vascular MNCD (ScVMNCD) according to Frisoni et al. (2002) and neuroimaging-supported probable diagnosis of vascular MNCD according to DSM-5; 16 subjects entered control group. At baseline, we detected lower BDNF levels in both MNCD groups, which was significant only in subjects with MNCD-AD. Moreover, plasma BDNF level of 21160 pg/mL showed high sensitivity (94%) to discriminate patients with MNCD-AD. Decreased plasma BDNF highly correlated with the severity of memory impairment and total MMSE score in MNCD-AD group. Escitalopram treatment in patients with MNCD-AD or ScVMNCD led to an increase of plasma BDNF concentrations and as a result to a decrease of cognitive, depressive, and anxiety symptom severity. In conclusion, plasma BDNF might be a reliable biomarker for the validation of MNCD-AD diagnosis and treatment efficacy.
Mild neurocognitive disorders (MNCDs) as an intermediate stage between normal cognitive aging and dementias, particularly Alzheimer’s disease (AD), have recently become a subject of an increasing scientific interest [
According to recent epidemiological data, the overall prevalence of MNCDs among individuals older than 55 is 15.7%, with single-domain amnestic, multiple-domain amnestic, and nonamnestic subtype prevalence of 6.4%, 3.7%, and 5.6%, respectively [
The investigation of neurobiological aspects of MNCDs might shed light on some pathogenetic mechanisms, which could become targets for management of MNCDs. The expression of growth factors, in particular brain-derived neurotrophic factor (BDNF), is one of them. In the majority of neurodegenerative and vascular dementias a reduction of BDNF concentration in the brain and concurrently in plasma [
Limited information is available for plasma/serum BDNF concentrations in patients with MNCDs. Although low levels of BDNF in serum [
Hence, the purpose of our study was to evaluate plasma BDNF concentrations in patients with the main etiological types of MNCDs and to determine whether the assessment of plasma BDNF level could improve the diagnostics of MNCD-AD and ScVMNCD. We also aimed to study the dynamics of plasma BDNF in MNCD-AD/ScVMNCD patients after escitalopram treatment. We chose escitalopram taking into account the evidence about its stimulation of BDNF expression [
59 persons over 65 years were enrolled in the study. 21 patients met the diagnostic criteria for possible MNCD-AD according to DSM-5 [
Clinical protocol included the following: (1) collection of anamnestic data; (2) neuropsychological testing: MMSE [
BDNF levels were measured by the ELISA method (kit supplied by R&D Systems, Inc., Minneapolis, USA), according to the manufacturer’s protocol using the immunoassay photometer “ImmunoChem-2100” (USA). Clinical studies and the evaluation of plasma BDNF concentrations were performed twice: at baseline in all enrolled persons and after 2 months of escitalopram treatment (daily dosage: 10 mg) in 20 patients with MNCDs. This dosage was prescribed in accordance with the FDA instruction for using the drug in elderly patients.
Statistical analyses were conducted using “STATISTICA 6.0” for Windows (StatSoft Inc., USA) v.6.1 and SPSS (version 19.0, SPSS Inc., Chicago, USA). The results were given as percentage, median, and interquartile range or mean and standard deviation, depending on the data distribution. The statistical significance of between-group comparisons was determined using parametric and nonparametric criteria when appropriate (Kruskal-Wallis test with subsequent multiple comparisons, Mann-Whitney test, Wilcoxon test,
The main demographic, cognitive, neuropsychiatric, neurological, and functional features of the comparison groups are summarized in Table
Main demographic, cognitive, psychiatric, neurological, neuropsychiatric, and functional characteristics in comparison groups.
Variables | Comparison groups |
|
||||
---|---|---|---|---|---|---|
WCI ( |
MNCD-AD ( |
ScVMNCD ( |
WCI versus MNCD-AD | WCI versus ScVMNCD | MNCD-AD versus ScVMNCD | |
Age (years) | 72.06 ± 5.25 | 73.90 ± 5.52 | 74.14 ± 5.69 | 0.478 |
||
Gender, male/female | 4/12 | 8/13 | 9/13 | 0.399 |
0.307 |
0.850 |
Education (years) | 12.88 ± 2.70 | 11.67 ± 4.51 | 12.59 ± 3.83 | 0.594 |
||
MMSE, score | 28 (28-29) | 26 (26-27) | 26 (24–27) | 0.000008 | 0.000001 | 0.93 |
Delayed recall, TIME test, number of words (0–5) | 4 (4-5) | 2 (1-2) | 2 (2-3) | <0.00001 | 0.0001 | 0.34 |
Delayed recognition, TIME test, number of words (0–10) | 9 (8–10) | 5 (5–7) | 8 (7–9) | <0.00001 | 0.04 | 0.0004 |
Clock drawing test, part I, score (0–10) | 9 (9-10) | 9 (9-10) | 7.5 (7-8) | 0.46 | 0.00001 | 0.0019 |
Verbal fluency, number of words during 3 min. | 19 (15–25.5) | 15 (12–21) | 13.5 (10–17) | 0.51 | 0.007 | 0.22 |
Kinetic apraxia, Luria’s tests, score (0–3) | 0 (0-0) | 0 (0-0) | 2 (1–3) | 1.0 | 0.00001 | 0.00001 |
Visuospatial apraxia, Luria’s tests, score (0–3) | 0 (0-0) | 1 (0-1) | 0 (0-1) | 0.019 | 0.89 | 0.19 |
The severity of pseudobulbar syndrome, score (0–3) | 0 (0-1) | 0 (0-1) | 2 (1-2) | 1.0 | 0.00002 | 0.00001 |
POMA, score (0–28) | 27.5 (26–28) | 26 (25-26) | 17.5 (13–21) | 0.06 | <0.00001 | 0.00002 |
NPI, score (0–144) | 7 (4–10) | 11 (8–14) | 12.5 (8–17) | 0.13 | 0.025 | 1.0 |
Depression, NPI ( |
5 (31.3) | 13 (61.9) | 19 (86.4) | 0.07 |
<0.0001 |
0.07 |
Anxiety, NPI ( |
15 (93.8) | 19 (90.5) | 19 (86.4) | 0.71 |
0.72 |
0.67 |
Apathy, NPI ( |
0 (0) | 9 (42.9) | 20 (90.9) | 0.003 |
<0.0001 |
0.001 |
Irritability, NPI ( |
9 (56.3) | 14 (66.7) | 10 (45.5) | 0.52 |
0.51 |
0.16 |
Sleep, NPI ( |
12 (75.0) | 19 (90.5) | 19 (86.4) | 0.21 |
0.37 |
0.73 |
BADL, score (0–60) | 0 (0-0) | 1 (1-1) | 2 (1-2) | 0.0001 | <0.00001 | 0.15 |
Mean ± standard deviation/median (lower/upper quartile) values are presented.
WCI patients without cognitive impairment.
MNCD-AD patients with mild neurocognitive disorder due to Alzheimer’s disease.
ScVMNCD patients with mild subcortical vascular neurocognitive disorder.
Nonparametric ANOVA for multiple comparisons if not otherwise specified.
The analysis revealed some significant clinical features that reliably distinguished patients with MNCD-AD and SсVMNCD. The most prominent feature of MNCD-AD patients was amnestic syndrome, which was significantly more severe in comparison with WCI. The severity of impairment of spontaneous delayed recall of five nouns was comparable in MNCD-AD and ScVMNCD groups. However, patients with MNCD-AD did not significantly improve the results of recall after cues. Subjects with MNCD-AD also had mild visuospatial apraxia when performing Luria’s tests. This symptom differentiated MNCD-AD patients from controls and those with ScVMNCD, who did not have any visuospatial disturbances. We also found no significant neurological disturbances in MNCD-AD group. Similarly, patients with ScVMNCD showed distinct spontaneous delayed recall impairment, but they had significantly higher rates of cued verbal recognition than those with MNCD-AD. In ScVMNCD group, we revealed pronounced executive and neurological disturbances in contrast to the WCI and MNCD-AD subjects. The executive dysfunction manifested in performing clock drawing test (part I), verbal fluency test, and Luria’s tests for kinetic apraxia. Moreover, there were mild to moderate pseudobulbar signs and frontal lobe gait disturbances (decreased POMA score) in ScVMNCD patients. The total score of psychopathological impairments (NPI) of ScVMNCD group was significantly higher than for WCI group, whereas the NPI score between MNCD-AD and WCI groups differed only at the level of trend. Predominant psychopathological symptoms in WCI group were anxiety (93.8%), insomnia (75.0%), and irritability/emotional instability (56.3%). In MNCD-AD group we observed anxiety and sleep disturbances (by 90.5%), irritability/emotional instability (66.7%), and depression (61.9%), whereas in ScVMNCD group we observed apathy (90.9%), depression, anxiety, and sleep disturbances (by 86.4%).
On average, MNCD patients had minimal disruption in everyday activities by BADL scale; however, it was significantly different from controls.
Table
Plasma BDNF levels (mean ± SD) in comparison groups at baseline, pg/mL.
Comparison groups |
|
||||
---|---|---|---|---|---|
WCI ( |
MNCD-AD |
ScVMNCD |
WCI versus MNCD-AD | WCI versus ScVMNCD | MNCD-AD versus ScVMNCD |
31581.5 ± 8092.2 | 19950.7 ± 9678.8 | 25939.6 ± 10410.5 | 0.0025 | 0.21 | 0.13 |
Next, we determined possible correlations between plasma BDNF concentrations and distinguishing clinical features of MNCD-AD and ScVMNCD groups described above (Table
Spearman’s/Pearson’s correlations between clinical, demographic, and cognitive variables and plasma BDNF concentrations in MNCD-AD and ScVMNCD groups.
Clinical variables | Spearman’s/Pearson’s correlation coefficient | |
---|---|---|
MNCD-AD group | ScVMNCD group | |
Age | −0.09 | −0.28 |
MMSE, total score | 0.49 |
0.18 |
Delayed recall, TIME test | 0.72 |
0.41 |
Clock drawing test, part I, score | 0.21 | 0.21 |
Verbal fluency, number of words | 0.32 | 0.09 |
Kinetic apraxia, Luria’s tests, score | −0.16 | −0.35 |
Visuospatial apraxia, Luria’s tests, score | −0.34 |
−0.33 |
NPI, total score | −0.22 | −0.12 |
NPI, the severity of depression, score | −0.17 | −0.22 |
BADL, score | −0.58 |
−0.25 |
We found highly significant positive correlations between plasma BDNF levels and the score of spontaneous delayed recall (TIME test, five unrelated nouns) in MNCD-AD group, as well as moderate positive correlations between plasma BDNF levels and general cognitive functioning (MMSE total score). Plasma BDNF concentrations negatively associated with visuospatial apraxia and impairment in daily living activities (BADL score).
As for the ScVMNCD group, plasma BDNF concentrations significantly correlated only with spontaneous word recall and visuospatial deficit. These relationships were less significant in comparison with the MNCD-AD group. We observed no significant correlations between plasma BDNF and psychopathological impairment or patient’s age in both MNCD groups. Hence, the reduction of BDNF levels predominantly influenced the severity of the amnestic syndrome in MNCD-AD patients and consequently the overall severity of cognitive and daily living impairment.
The discriminating ability of plasma BDNF for the MNCD-AD/ScVMNCD and control was determined by the AUC analysis. Figure
ROC curves for plasma BDNF to diagnose MNCD-AD (a) and ScVMNCD (b). AUCs (95% confidence intervals) are 0.866 (0.733–0.999) for MNCD-AD (
After two months of escitalopram treatment, 20 patients (10 in each MNCD group) underwent repeated clinical examination and evaluation of plasma BDNF concentrations. The results are shown in Table
The dynamics of plasma BDNF concentrations and main clinical characteristics in MNCD patients taking escitalopram.
Variables | MNCD-AD group ( |
ScVMNCD group ( |
||||
---|---|---|---|---|---|---|
Before treatment | After treatment |
|
Before treatment | After treatment |
| |
Plasma BDNF concentrations, pg/mL | 20660.4 ± 12774.6 | 26356.0 ± 8309.1 | 0.15 |
26652.4 ± 12435.7 | 30066.0 ± 10796.4 | 0.27 |
MMSE, score | 26.5 (26-27) | 28 (28-29) | 0.005 | 25 (24–26) | 27 (26–28) | 0.005 |
Delayed recall, TIME test, number of words (0–5) | 2 (1-2) | 3.5 (3–5) | 0.02 | 2.5 (2-3) | 3 (1–4) | 0.83 |
Verbal fluency, number of words during 3 min. | 19.5 (15–25) | 17.5 (12–24) | 0.62 | 13 (10–17) | 17 (10–19) | 0.20 |
Clock drawing test, part I, score (0–10) | 9 (8–10) | 9 (8–10) | 0.55 | 7 (7-8) | 7.5 (6–9) | 0.23 |
Kinetic apraxia, Luria’s tests, score (0–3) | 0 (0-1) | 0 (0-0) | 0.22 | 2 (1–3) | 2 (1-2) | 0.48 |
NPI, total score (0–144) | 11.5 (9–14) | 6.5 (6–8) | 0.0078 | 16.5 (9–18) | 10.5 (7–13) | 0.011 |
POMA, score (0–28) | 25 (24–26) | 25 (24–26) | 1.0 | 15.5 (13–21) | 16 (15–21) | 0.043 |
BADL, score (0–60) | 1 (1-1) | 1 (0-1) | 0.11 | 2 (1-2) | 2 (1-2) | 0.98 |
Data are presented as mean ± standard deviation/median (lower/upper quartile).
The intake of escitalopram increased plasma BDNF concentrations in both MNCD groups; nevertheless, those differences did not reach statistical significance. Simultaneously, escitalopram treatment improved clinical parameters in patients with MNCD-AD and ScVMNCD. Thus, we observed a significant increase in general level of cognitive functioning (MMSE total score) and a decrease in the frequency and severity of neuropsychiatric symptoms (mainly depression and anxiety) by NPI scale. We detected significant improvement in cognitive functioning in the MNCD-AD group, predominantly in memory domain (spontaneous delayed recall). At the same time, patients with ScVMNCD mainly improved executive functions (verbal fluency test, clock drawing test, and kinetic praxis), although the results were not statistically significant. The overall score of impairment in daily living activities (BADL score) decreased at the level of trend in both MNCD groups.
In this study, we found a decrease in plasma BDNF concentrations in patients with the main etiological types of MNCDs. The decrease was significant in patients with MNCD-AD and correlated with the severity of amnestic syndrome. Plasma BDNF was found to be highly sensitive in distinguishing MNCD-AD patients from WCI subjects. After escitalopram treatment, we observed an increase in plasma BDNF levels in both MNCD groups and simultaneous improvement of distinguishing clinical features of MNCD-AD and ScVMNCD. Moreover, we found a good accuracy of plasma BDNF in discriminating patients with MNCD-AD and persons WCI. Our results are consistent with findings for amnestic MNCD obtained by Hwang and colleagues [
As it was shown, brain neurodegenerative processes at the early AD stages could affect neurons and glial cells in the main areas of BDNF synthesis, such as the hippocampus, amygdala, and neocortex [
The link between decreased BDNF concentrations and the severity of amnestic syndrome indicates a key role of BDNF in memory impairment among MNCD-AD patients. Jovanovic et al. reported that BDNF increased the opening of NMDA receptors and enhanced expression of AMPA receptors of the postsynaptic membrane, resulting in facilitating of long-term potentiation (LTP) [
The ability of plasma BDNF level to discriminate patients with MNCD-AD makes it a candidate biomarker for early identification, monitoring, and intervention assessment in such an etiological type of MNCDs.
Our study demonstrated that escitalopram treatment (10 mg per day) led to an increase in plasma BDNF levels in both MNCD-AD and ScVMNCD groups. The increase of plasma BDNF levels correlated with a significant improvement in cognitive functioning and reduction of neuropsychiatric symptoms (predominantly depression and anxiety) in both MNCD groups as well as with significant amelioration in memory performance in MNCD-AD patients and uncertain improvement in executive functions in ScVMNCD persons. According to Ladea and Bran, treatment with escitalopram can significantly reduce the severity of depressive symptoms and simultaneously increase serum BDNF concentrations in elderly patients [
This study has some limitations. As a preliminary study, our sample size was relatively small. Another limitation is that the ELISA kit we used quantified the total BDNF concentrations without distinction between pro-BDNF and mature BDNF variants. Furthermore, our findings are based on the concept that plasma BDNF level is associated with alterations in the brain [
In conclusion, our study demonstrated a decrease of plasma BDNF concentrations in patients with the main MNCDs’ etiological types, which was more significant in subjects with MNCD-AD. In patients with MNCD-AD, the reduction of plasma BDNF concentrations was predominantly associated with the memory process impairment. Plasma BDNF levels showed high sensitivity for detecting MNCD-AD. Escitalopram treatment in patients with MNCD-AD and ScVMNCD led to an increase in plasma BDNF concentrations and, as a result, to a decrease of cognitive deficits, as well as to a decrease of depressive and anxiety symptom severity. Plasma BDNF concentrations might be used as a reliable biomarker for the validation of MNCD-AD diagnosis and assessment of treatment efficacy.
The authors have no conflict of interests to declare.