Circulating brain‐derived neurotrophic factor levels and heart failure: A systematic review and meta‐analysis

Abstract

Aims

Biomarkers are paramount for managing heart failure (HF) patients as prognostic and therapeutic efficacy index tools. Systemic levels of brain‐derived neurotrophic factor (BDNF) can add to the HF biomarker scenario, allowing for potentiated efficacy in diagnosis, prognostic stratification, and prediction of patient response to a given therapeutic intervention because BDNF is one of the primary rulers of myocardial function. Yet, whether BDNF is a reliable clinical biomarker awaits clinical validation. Hence, we aimed to answer this relevant question via a systematic review and meta‐analysis of existing studies.

Methods and results

International databases, including PubMed, Scopus, Embase, and the Web of Science, were comprehensively searched for studies assessing BDNF levels in patients with HF versus non‐HF controls or as a prognostic factor for HF complications. Data were extracted and analysed by random‐effect meta‐analysis. Standardized mean difference (SMD) and 95% confidence intervals (CIs) were computed to pool the results of studies. We included 11 studies in the final review, among which six underwent meta‐analysis. These studies analysed 1420 HF patients, with a mean age of 65.4 ± 11.2 years. Meta‐analysis revealed that patients with HF had significantly lower circulating BDNF levels than healthy controls (SMD −2.47, 95% CI −4.39 to −0.54, P‐value = 0.01). Moreover, patients with higher New York Heart Association functional classification had lower levels of BDNF. Adverse clinical outcomes such as all‐cause mortality and HF rehospitalization were also associated with lower levels of BDNF in individual studies.

Conclusions

BDNF levels are decreased in patients with HF. Most importantly, we observed an association between lower BDNF levels and poor prognosis in patients with HF. Our study supports BDNF as an easy‐to‐dose diagnostic and prognostic biomarker to be implemented in clinical practice for HF. Further studies are warranted to address this ability specifically.

Introduction

Heart failure (HF) is a chronic progressive syndrome and a global health problem involving millions worldwide.^1,2 This disorder represents the end-stage condition of several cardiovascular (CV) diseases and non-CV diseases, including, and not limited to, myocardial infarction (MI), valvular diseases, hypertension, arrhythmias, and diabetes culminating in adverse cardiac remodelling, decline in contractile function, and neurohormonal alterations.^3,4 At the molecular level, altered crosstalk between the brain and the heart appears central to these negative features. Indeed, changes in how the autonomic nervous system (ANS) modulates cardiac function are pivotal to this miscommunication.^3,5,6 Certainly, in HF, it has been recognized that the abnormal activation of the sympathetic nervous system (SNS) leads to further worsening of the disease.

Hence, alongside the well-recognized biomarkers (e.g., natriuretic peptides, troponins, and galectin-3) proposed in managing HF that fulfil the criteria of accuracy, quick, easy, high sensitivity and specificity, and reproducible dosing and reflect the pathophysiological processes of HF other molecules involved in ANS activation can be potentially employed in routine practice.^3,7,8

Brain-derived neurotrophic factor (BDNF) is an on-demand neurotrophin (NT) produced and released locally and systemically by neurons, whose expression is also regulated by adrenergic receptors.⁹ In the brain, BDNF exerts important pleiotropic activities, promoting cell growth and connectivity by binding to its specific receptor, tyrosine receptor kinase B (TrkB), in physiological conditions and in response to stress conditions.¹⁰ However, the neuronal district is not the sole contributor to circulating BDNF levels.¹¹ Indeed, studies determined that the BDNF and its receptors are expressed by other cell types, including cardiac myocytes,^12–14 endothelial cells,¹⁵ and vascular smooth muscle cells,¹⁶ contributing to the integrity of the cardiovascular system homeostasis. In this regard, BDNF exerts trophic actions on peripheral nerves in the cardiovascular system, but it can also directly control cardiac and vascular development and functioning. Moreover, BDNF regulates cardiac bioenergetics by modulating the transcription factor Yin Yang 1.¹⁷ In the setting of cardiac disorders, reduced levels of circulating BDNF are associated with adverse cardiac remodelling and more elevated levels of NTproBNP.¹⁸ In line with these data, Barman and colleagues¹⁹ showed that a reduction in peripheral BDNF levels was associated with death and rehospitalization in patients with HF with reduced ejection fraction (HFrEF). Furthermore, these authors demonstrated that BDNF levels resulted lower than those observed in age-and sex-matched healthy controls. At molecular levels, studies in mice have shown that impaired BDNF/TrkB system contributes to the onset and maintenance of chronic post-ischemic left ventricular adverse remodelling and decompensation. Of note, BDNF levels decline in the cardiovascular system of aged individuals.⁶ Thus, BDNF may represent a novel potential easy-to-dose HF diagnostic and prognostic biomarker.

Here, we explored the connection between BDNF and HF in this study. Of note, the influence of BDNF was also investigated within the complications of HF, including all-cause mortality, HF rehospitalization, and cardiac events.

Methods

Search strategy

This study was conducted based on the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guideline²⁰ for study design, search approach, and reporting. A comprehensive systematic search was carried out in PubMed, Web of Science, Embase, and SCOPUS up to 27 January 2023, to identify relevant data, without any filter or restriction. The applied search terms used in search queries were: ‘Brain-Derived Neurotropic Factor’ OR ‘BNDF’ AND ‘heart failure’ OR ‘myocardial failure’ OR ‘heart decompensation’ OR ‘left ventricular dysfunction’ as well as other commensurate terms. The detailed search strategy is represented in Table S1.

Study selection and screening

Studies that met the eligibility criteria were included (1) clinical studies that identified the serum level of BDNF in HF patients and compared them with a control group and (2) studies that investigated BNDF levels in different stages of HF. Excluded studies were categorized into one of the following groups: (1) case reports, case series, and other studies without a control group; (2) conference abstracts, letters, commentaries, reviews, and in vitro studies; and (3) animal studies; and (4) non-English published articles.

After the elimination of duplicate reports, the remaining studies' titles and abstracts were screened by two researchers (A. H. B. and A. K.) separately to assess their appropriateness for inclusion. Subsequently, the previously selected papers were evaluated for eligibility using full text. If there were discrepancies between the two reviewers' opinions, the third researcher (Q. M.), as a referee, concluded. We also noticed the reference list of the finally included articles to find further qualified studies.

Data extraction and quality assessment

A pre-defined list was prepared owing to the extraction of the required data from the categorized final articles exhaustively which contained: (1) the first author's name, publication year, and country of investigation; (2) inclusion criteria for the heart failure (population) and the control group; (3) the demographic characteristics of the population (sample size, mean age ± standard deviation, and gender distribution in each case and control group); and (4) main findings. All of the data mentioned above were obtained from each article by two authors (A. E. E. K. and A. K.) independently.

The ‘Newcastle–Ottawa Quality Assessment Scale’ (NOS) for nonrandomized studies²¹ was utilized to evaluate the risk of bias in included articles. This process was conducted by two investigators independently (A. K. and A. H. B.) in addition to a third person (Q. M.) at the place of the decider in the case of any dispute about quality assessment. Selection, comparability, and outcome/exposure are three quality evaluation parameters assigned a maximum of 4 points, 2 points, and 3 points for each article, respectively. The highest quality study is reflected by the score of 9, as stated.

Statistical analysis

We assessed standardized mean difference (SMD) with its 95% confidence interval (CI) to compare BDNF levels between HF patients and the control group or within different stages of HF patients as there were adequate studies. All the statistical analyses were conducted using Stata Statistical Software (version 17.0, Stata Corp), and statistically significant effects have been determined as effects with P-values < 0.05. The statistical heterogeneity of studies was calculated with the Q and quantified by Higgin's I². A value of I² of ≤25% represents low heterogenicity, 26–75% represents moderate heterogenicity, and >75% represents a high heterogenicity.²² Due to high heterogenicity, we preferred to utilize the random-effect model (DerSimonian and Laird) for the meta-analyses. A sensitivity analysis was carried out by eliminating each study (Leave-one-out method) and examining its effect on pooled estimates. In order to identify the source of heterogenicity, meta-regression of the publication year and mean age, in addition to subgroup analysis, was also performed. As well as Egger's²³ and Begg's²⁴ statistical tests, we identified publication bias utilizing a visual assessment of funnel plots.

Results

Literature search and studies included

The initial search resulted in 723 total studies comprised 72 from PubMed, 342 from Embase, 101 from the Web of Science, and 208 from Scopus. After removing duplicated studies (n = 281), 442 studies remained to be screened (Table S1). In the title/abstract screening process, 35 studies remained for full-text assessment. Finally, 11 studies remained to be included in the systematic review. Details of the search and screening process are shown in Figure 1 while Table 1 provides the baseline characteristics of the 11 studies that were included in this study.^19,25–34 A total of 1904 individuals were assessed in these studies, among which 1420 were patients with HF. The mean age of these patients was 65.4 ± 11.2 years and 68.6% were male. Six studies were conducted in Japan,^{26–29,31,32} while the remaining 5 were performed in China.^25,33,34 All the studies analysed had high qualities, measured by NOS (Table 2).

[IMAGE OMITTED. SEE PDF]

Table 1 Baseline characteristics and main findings of the included studies

Author	Year	Location	Population (N)	Controls (N)	Mean age (years)	Male (%)	Main findings
Barman et al.	2019	Turkey	Patients with LVEF ≤35%, 52 were classified as mildly symptomatic (NYHA I–II) and 108 as severely symptomatic (NYHA III) (N = 160)	Individuals (N = 50)	67.6 ± 11.4	58	Low serum BDNF levels were associated with death and rehospitalization in HF patients, suggesting that these levels can be useful prognostic biomarkers. Patients with LVEF < 35% had higher NT-proBNP concentration and lower BDNF than the control group without cardiac disease.
Fang et al.	2022	China	HF > half a year, in line with the diagnostic criteria for CHF with HF treatment for 1 ≥ month after the study (N = 82)	Healthy patients (N = 78)	64.9 ± 1.4	54.88	MiR-182-5p expression is increased and BDNF level is decreased in CHF patients. MiR-182-5p combined with BDNF can assist in the diagnosis of CHF and predict a poor prognosis.
Fukushima et al.	2013	Japan	HF patients (N = 43)	Healthy individuals (N = 27)	57.4 ± 14.8	76.75	Serum BDNF levels correlated with exercise capacity in HF, suggesting that serum BDNF could be a useful marker for predicting exercise capacity and the long-term outcomes of HF patients.
Fukushima et al.	2015	Japan	HF patients, NYHA (I/II/III) 5/39/14 (N = 58)	NA	59.2 ± 13.7	75.9	Decreased serum BDNF levels were significantly associated with adverse outcomes in HF patients, suggesting that these levels can be a useful prognostic biomarker.
Kadowaki et al.	2016	Japan	CHF patients (patients with stable HF and patients with decompensated HF) (N = 134)	Age-matched subjects without signs of significant heart disease (N = 23)	71 ± 13	59.7	Low serum BDNF levels were found in patients with CHF (advancing NYHA functional class), and these levels were found to be independently associated with an increased risk of cardiac events.
Nakano et al.	2020	Japan	CHF patients NYHA functional class I–III with a history of ≥hospitalizations due to worsening HF and ≥1 month post-discharge without worsening (N = 60)	Age-matched controls (N = 29)	63 ± 12	58.33	The serum BDNF level was significantly lower in the HF patients compared with the controls (24.9 ± 0.9 vs. 28.6 ± 1.3). Analysis revealed that BDNF was independently associated with muscle strength (β-coefficient = 2.80, 95% CI 1.89 to 11.8). Serum BDNF levels were associated with skeletal muscle function, but not with muscle mass. Speculations that BDNF is secreted by muscle activity, regulates mitochondrial function in the skeletal muscle, and defines exercise endurance.
Pytka et al.	2022	Poland	Ambulatory patients with a compensated HF and LVEF below 50% (N = 361)	N/A	63.8 ± 9.9	84.5	Compared with people with higher levels of BDNF, HF patients with LVEF < 50% and lower serum BDNF concentration have more advanced cardiac remodelling and dysfunction.
Shibata et al.	2018	Japan	Hospitalized HF patients with cardiac rehabilitation (N = 94)	N/A	68 ± 14.5	63.8	In the low BDNF and low peak VO2 group, the frequency of rehospitalization within half a year after discharge was higher than that in other groups. Serum BDNF levels at discharge may be a useful biomarker for the prognosis in HF patients. Furthermore, combining BDNF and exercise tolerance may be useful for predicting early cardiac events.
Takashio et al.	2015	Japan	HF patients (N = 242)	Subjects without HF-age and gender matched (N = 80)	71 ± 12	65.7	The plasma BDNF levels were significantly lower in HF patients than those without HF (3712 pg/mL [2124 to 6180] vs. 7247 pg/mL [5388 to 9255], P < 0.001) and lower in HF patients with NYHA III than class I (P < 0.01) and class II (P < 0.001). Log BDNF level was found to be a significant predictor of the occurrence of HF in multivariate logistic regression analysis (odds ratio 0.82; 95% confidence interval 0.76 to 0.91, P < 0.001). In conclusion, plasma BDNF levels were lower in HF patients and associated with HF severity. BDNF could be a potentially clinically useful biomarker of HF reflecting possible cardio-neuronal link-age.
Wu et al.	2019	China	Patients clinically diagnosed cardiac function II–IV; and AHF (N = 78)	Patients who had cardiac function I (without AHF) (N = 82)	65.8 ± 8.3	69.23	CA125 and BDNF play a role in the occurrence and development of AMI with AHF. As the disease progresses, serum CA125 and BDNF levels increase and are positively correlated with the severity of AMI combined with AHF. Both CA125 and BDNF have a good diagnostic value for AHF, and their combined detection can improve the sensitivity of an AHF diagnosis.
Xie et al.	2023	China	HF patients (N = 108)	Healthy individuals (N = 115)	53.7 ± 9.9	61.1	The serum sST2, CTnI, and BUN/Cr were high in HF patients and BDNF low. Medical practitioners should pay attention to the risk factors sST2, CTnI, BUN/Cr and a higher BNDF indicates better condition in HF patients. The BDNF and LVEF were negatively correlated with cardiac function, with r = −0.43, P < 0.001 and r = −0.39, P < 0.001, respectively. The logistic regression analysis with heart failure as the dependent variable revealed that sST2, CTnI, BUN, Cr, and BUN/Cr were the risk variables for heart failure and BDNF and LVEF were the protective factors.

Table 2 Quality assessment of included studies based on the Newcastle-Ottawa Scale (NOS)

Study	Selection	Comparability	Outcome	Overall score
Representation	Sample size	Non-respondents	Exposure	Outcome	Statistical test
Barman et al. (2019)	*	*	*	**	**	**	*	10
Fang et al. (2022)	*	*	*	**	**	**	*	10
Fukushima et al. (2013)	*	*	*	**	*	**	*	9
Fukushima et al. (2015)	*	*	*	**	-	**	*	8
Kadowaki et al. (2016)	*	*	*	**	*	**	*	9
Nakano et al. (2020)	*	*	*	**	*	**	*	9
Pytka et al. (2022)	*	*	*	**	-	**	*	8
Shibata et al. (2018)	*	*	*	**	-	**	*	8
Takashio et al. (2015)	*	*	*	**	**	**	*	10
Wu et al. (2019)	*	*	*	**	-	**	*	8
Xie et al. (2023)	*	*	*	**	-	**	*	8

Meta-analysis of brain-derived neurotrophic factor levels in patients with heart failure versus healthy controls

Six studies compared BDNF levels between HF patients and healthy controls and reported serum and/or plasma BDNF concentration.^{19,27–29,32,34} Random-effect meta-analysis of blood BDNF levels showed that patients with HF had significantly lower levels than non-HF cases (SMD −2.47, 95% CI −4.39 to −0.54, P-value = 0.01). However, this was associated with high heterogeneity (I²: 99.1%). The forest plot showing this meta-analysis of six studies is illustrated in Figure 2.

[IMAGE OMITTED. SEE PDF]

Meta-regression showed a negative relationship between the effect size observed and the publication year (slope −0.455, 95% CI −0.893 to −0.017, P-value = 0.042). Moreover, as described in Table 3, the mean age of patients and publication year accounted for 14.4% and 39% of the variance between the studies, respectively. The bubble plots for all meta-regression analyses are in Figures S1–S5. The funnel plot for assessing publication bias showed no apparent asymmetry (Figure S6). However, both Begg's and Egger's statistical tests were significant for publication bias (P-value = 0.024 and 0.022, respectively). The Leave-one-out method was used to assess each study's effect on the overall effect size. It was shown that removing Barman et al.¹⁹ resulted in an insignificant difference in BDNF levels (SMD −2.22, 95% CI −4.50 to 0.06, P-value = 0.056). The removal of other studies had no effect in terms of pooled estimate significance (Figure S7).

Table 3 Meta-regression of meta-analysis for BDNF levels in patients with heart failure versus controls

Moderator	No. of comparisons	Meta-regression	R2 analog (proportion of variance explained)
Heart failure	Control	Slope	95% confidence interval	P-value
Mean age (years)	747	324	0.187	−0.083	0.457	0.175	14.35%
Publication year	747	324	−0.455	−0.893	−0.017	0.042	39%
Male sex (percentage)	747	324	0.079	−0.238	0.397	0.623	0%
Sample size	747	324	−0.008	−0.032	0.016	0.506	0%
Left ventricular ejection fraction	747	324	0.020	−0.308	0.348	0.905	0%

Diagnostic ability of brain-derived neurotrophic factor in heart failure

The diagnostic role of BDNF for HF was assessed in four studies. Fang et al.²⁵ found that a cut-off of 23.70 ng/mL can diagnose chronic HF with an AUC of 0.829; however, BDNF in combination with miR-182-5p improves its diagnostic capacity (AUC 0.894). In line with these data, Takashio and coworkers³² reported that log BDNF levels significantly detected the presence of HF (adjusted OR 0.82, 95% CI 0.76 to 0.91, P-value<0.001). For their part, Wu and colleagues³³ evaluated BDNF levels in a cohort of 160 patients with acute myocardial infarction (AMI). These authors observed higher BDNF levels in patients with AMI compared with controls (P-value<0.001). Interestingly, these authors provided data suggesting BDNF as an appropriate marker to distinguish patients developing acute HF (AHF) from non-AHF ones (AUC 0.880, 95% CI 0.826 to 0.934). In addition, in these patients, Wu et al. also analysed the levels of the Antigen carbohydrate 125 (CA125), a promising diagnostic and prognostic biomarker of AHF,³⁵ and demonstrated that this biomarker enhanced BDNF's diagnostic ability (AUC 0.965, 95% CI 0.941 to 0.989). Finally, Xie et al.³⁴ showed that BDNF was protective against HF (OR 1.220, 95% CI 1.068 to 1.301, P-value = 0.001).

Brain-derived neurotrophic factor levels in heart failure patients according to functional capacity

Four studies evaluated BDNF levels in the New York Heart Association (NYHA) functional classification in patients with HF. Barman et al.¹⁹ found that BDNF levels were significantly lower in NYHA III compared with NYHA I-II (9.77 ng/mL vs. 10.22 ng/mL). Analogously, Kadowaki et al.²⁸ reported that BDNF levels were reduced in NYHA III and IV compared with NYHA II (P-value<0.05). Next, in line with these data, Takashio and colleagues³² demonstrated that BDNF levels decreased in NYHA class III patients (2563 pg/mL) compared with NYHA II (4052 pg/mL, P-value<0.001), NYHA I (4622 pg/mL, P-value = 0.01), and non-HF individuals (7247 pg/mL, P-value<0.001). Finally, Xie and coworkers³⁴ observed a significant reduction in BDNF levels in higher stages of NYHA (NYHA II 21.02 ± 2.01 ng/mL, NYHA III 18.23 ± 1.78, NYHA IV 14.02 ± 1.47, controls 32.01 ± 2.23).

Prognostic value of brain-derived neurotrophic factor in heart failure patients

Cardiac events (all cardiac mortality and heart failure rehospitalization)

Four studies investigated the prognostic role of BDNF in predicting cardiac events (i.e., all cardiac mortality and HF rehospitalization) in HF patients. In a median follow-up of 29.4 months, Barman and colleagues¹⁹ found that low levels of BDNF act as an independent indicator of all cardiac deaths (adjusted OR 0.170, 95% CI 0.054 to 0.533, P-value = 0.002). Indeed, these authors found an AUC of 0.837 [95% CI 0.735 to 0.938] in predicting all cardiac death in HF patients. In addition, they demonstrated that low levels of BDNF were an independent indicator of HF rehospitalization (adjusted OR 0.702, 95% CI 0.538 to 0.917, P-value = 0.010). For their part, Fukushima and coworkers²⁶ found that a reduction in BDNF levels was associated with higher adverse clinical outcomes (adjusted HR 0.41, 95% CI 0.20 to 0.84, P-value = 0.003) with an AUC of 0.798. Importantly, these authors observed that HF patients presenting low BDNF (<17.4 ng/mL) had significantly more adverse events compared with those with high BDNF (≥17.4 ng/mL; 66% vs. 16%, P-value<0.001).

Next, Kadowaki et al.²⁸ compared BDNF levels between HF patients with and without cardiac events demonstrating that BDNF levels proved to be lower in patients with cardiac events (i.e., cardiac death and HF rehospitalization) than those event-free individuals (12.5 ± 8.5 ng/mL vs. 16.1 ± 8.0 ng/mL, P-value = 0.0169). This study demonstrated that BDNF could predict adverse cardiac events with an AUC of 0.639 (cut-off = 12.4 ng/mL). In this regard, using a log-rank test these authors found greater cardiac events in those patients with low BDNF (≤12.4 ng/mL) compared with the group of patients with high BDNF (>12.4 ng/mL; P-value = 0.0005). In addition, they demonstrated that low BDNF (≤12.4 ng/mL) was an independent predictor of cardiovascular events in HF patients (adjusted HR 2.932, 95% CI 1.622 to 5.301, P-value = 0.0004).

Finally, in the study by Shibata et al.,³¹ HF patients were divided into groups of patients: high-BDNF (>19.22 ng/mL) and low-BDNF (≤19.22 ng/mL). This cut-off detected cardiac events with 57% sensitivity and 69% specificity with an AUC of 0.67 (P-value = 0.005). The low-BDNF group presented with a higher risk of cardiac events according to the log-rank test (P-value = 0.023) compared with the high-BDNF group. In addition, as demonstrated by this study, BDNF resulted in an independent predictor of cardiac events in HF patients (adjusted HR 0.956, 95% CI 0.911 to 0.999, P-value = 0.046).

All-cause mortality

Using a log-rank test, Takashio and colleagues³² demonstrated that in a follow-up of 663 days, patients with low BDNF (<3.71 ng/mL) had significantly higher all-cause mortality (P-value = 0.01). Moreover, as reported by these authors low-BDNF acts as an independent predictor of all-cause mortality in HF patients (adjusted HR 2.22, 95% CI 1.03 to 4.82, P-value = 0.04).

Effect of acute decompensated heart failure treatment on brain-derived neurotrophic factor levels

Takashio et al.³² found that treatment of acute decompensated HF (ADHF) significantly increased BDNF levels in these patients (admission: 2749 pg/mL [IQR 1380 to 4161] vs. discharge: 4194 pg/mL [2356 to 6916], P-value = 0.003).

Discussion

In this study, we performed a comprehensive systematic review and meta-analysis on a total of 1904 individuals assessed in n = 11 studies to investigate the association between BDNF and HF. Interestingly, the principal findings of this study are that (1) BDNF circulating levels in patients with HF are significantly impaired compared with healthy individuals; (2) BDNF levels can stratify HF patients according to NYHA functional capacity; (3) in HF patients, lower circulating levels of BDNF could be a prognostic factor in adverse outcomes such as cardiac events, rehospitalization, and mortality. It should be noted that removing Barman et al.¹⁹ resulted in an insignificant difference in BDNF levels between HF patients and controls.

BDNF is a member of the NT family and is one of the most abundant forms in the brain, altered in several conditions.^36–40 Like the other NTs, the BNDF plays an essential role in the early stages of development and throughout life, as it regulates differentiation, growth, and maintenance of the nervous system.⁴¹ For this reason, numerous studies have investigated and demonstrated that lower circulating BDNF levels are associated with central nervous system complications and psychiatric disorders like depression,⁴² anxiety,⁴³ Alzheimer's disease,⁴⁴ and Parkinson's disease,⁴⁵ demonstrating the potential role of this NT as a biomarker.

However, BDNF and his TrkB receptor have also been detected in non-neuronal cells, including those of the cardiovascular system (e.g., endothelial and vascular smooth muscle cells and cardiomyocytes). BDNF signalling is required in the heart for proper myocardial contraction and relaxation.⁴⁶ Notably, in response to ischemic stress, like MI, the circulating BDNF levels are increased,⁴⁷ and part of the BDNF secreted is directly produced by cardiomyocytes in response to increased adrenergic stimulation, typically observed after MI.¹⁴

BDNF protects against ischemic injury and prevents cardiac remodelling following MI.^14,47,48 However, when chronic LV decompensation ensues, BDNF levels are reduced^14,47,49 likely contributing to impaired innervation, angiogenesis, and cardiomyocyte function.¹⁴ Of relevance, reinstituting BDNF signalling arrests post-ischemic chronic HF progression in mice.^14,49 Indeed, as recently demonstrated by Li et al.,¹² the selective genetic ablation of BDNF in cardiomyocytes resulted in impaired cardiac function, hypertrophy of cardiomyocytes, myocardial degeneration, and inflammation leading to HF.¹² Based on this premise, it appears clear that a loss in BDNF underscores chronic post-ischemic HF, further supporting our contention that BDNF could serve as an additional significant biomarker in HF patients, especially the post-ischemic ones. However, it is worth noting that impairment in (both tissue and plasma/serum) BDNF levels has also been associated with several pathological HF-leading disorders. For instance, it has been reported that circulating BDNF level reduction may be involved in coronary atherosclerosis pathogenesis, and decreased levels of this factor contribute to acute coronary syndrome and mortality.⁵⁰ Further, studies have revealed that BDNF levels are significantly lower in diabetes mellitus patients compared with healthy individuals⁵¹ and these patients have a two-fold higher risk of HF development.^52,53 While other studies have demonstrated a relationship between anxiety and depression in HF progression and development,⁵⁴ and individuals with such disorders displayed lower plasma BDNF levels.⁴³ Finally, as BDNF is released from skeletal muscles, a loss in muscle mass usually observed in HF patients may also significantly contribute to the reduction in peripheral BDNF secretion.^27,55 Therefore, assessing BDNF levels may represent an important diagnostic and prognostic tool for cardiovascular disease and HF and this is a central finding of our study.

Our findings suggest that there is an association between BDNF levels and HF while lower BDNF levels could also be associated with worse prognosis in HF. However, larger diagnostic and prognostic studies are needed to confirm these findings and pave the way for this novel biomarker to be suggested in HF guidelines. The fact that BDNF is affected by several conditions and diseases should also be taken into consideration when interpreting the results. Hence, our results are not conclusive and several other studies are needed to claim if BDNF is a practically useful biomarker for HF. Currently, there are several well-known markers measured routinely in HF including proBNP and mid-regional proadrenomedullin (MR-proADM). However, recent studies found some limitations in using these biomarkers in patients with preserved LVEF. They found NT-proBNP is a ‘less perfect’ biomarker in HFpEF compared with reduced LVEF.⁵⁶ A study found no additional information from MR-proADM compared with NT-proBNP.⁵⁷ Another study evaluated the role of CA125 and BDNF in predicting HF after AMI and found promising promising results,³³ showing that the combination of CA125 and BDNF has a higher predictive ability in acute HF.^33,58 The diagnostic and prognostic significance of BDNF in preserved LVEF HF is still unknown. As known markers have some limitations in these patients, comparative studies aiming at determining the relative value of BDNF should be conducted.

Limitations

This study has three main limitations. The first limitation is the small number of studies included in the analysis. For each of the analyses and sections, there was a low number of studies included and this can lead to a decrease in the power of our analyses. Importantly, our results are not conclusive and a definite conclusion could not be made. Hence, to better understand the relationship between BDNF and HF, we need larger sample sizes and more specific clinical characterization in the future. The second limitation is that our meta-analysis presents high heterogeneity (I²: 99.1%). This could stem from the minor differences in the populations assessed in each of the meta-analysed studies. Moreover, we found that 53.37% of heterogeneity can be attributed to mean age and publication year in our meta-regression analysis. Therefore, robustness and similar studies are needed to solidify these results with a lower overall heterogeneity. Third, most of the studies were conducted in China and Japan in eastern Asia, which can limit the generalizability of these findings. Hence, further studies in other regions should be conducted to better confirm these findings. In addition, sensitivity and specificity values for association between BDNF and several variables were quite low, limiting the diagnostic and prognostic role of BDNF. Finally, we could not perform subgroup analyses due to the limited number of studies that estimated plasma vs serum BDNF levels.

Conclusions

BDNF circulating levels are a promising and valuable clinical diagnostic biomarker for HF, especially the post-ischemic one. Also, as BDNF could show the severity and risk of complications and mortality, it could be used as a prognostic indicator. Among the well-recognized biomarkers used in clinical practice, BDNF could play a dual role, reflecting both the pathophysiological processes of myocardial stress, cardiac fibrosis, and systemic inflammation and explicitly assessing SNS derangement typical of HF. For this reason, further studies and correlation with some of the parameters primarily involved in the alteration of circulating BDNF levels like SNS activity and catecholamines levels, depression, skeletal mass, and others should be performed to identify a proper cut-off level for BDNF in HF and to aid clinicians in appropriate diagnosis and monitoring of patients.

Acknowledgements

We are very thankful to Dr. Nazareno Paolocci (JHU) for his critical manuscript reading and suggestions.

Conflict of interest

The authors declare that they have no competing interests.

References

Savarese G, Lund LH. Global public health burden of heart failure. Card Fail Rev 2017;3:7‐11. doi:10.15420/cfr.2016:25:2

Lippi G, Sanchis‐Gomar F. Global epidemiology and future trends of heart failure. AME Med J 2020;5: doi:10.21037/amj.2020.03.03

Bencivenga L, Palaia ME, Sepe I, Gambino G, Komici K, Cannavo A, et al. Why do we not assess sympathetic nervous system activity in heart failure management: might GRK2 serve as a new biomarker? Cell 2021;10: [eLocator: 457].

Khalaji A, Behnoush AH, Khanmohammadi S, Ghanbari Mardasi K, Sharifkashani S, Sahebkar A, et al. Triglyceride‐glucose index and heart failure: a systematic review and meta‐analysis. Cardiovasc Diabetol 2023;22:244.

Florea VG, Cohn JN. The autonomic nervous system and heart failure. Circ Res 2014;114:1815‐1826.

Elia A, Cannavo A, Gambino G, Cimini M, Ferrara N, Kishore R, et al. Aging is associated with cardiac autonomic nerve fiber depletion and reduced cardiac and circulating BDNF levels. J Geriatr Cardiol 2021;18:549‐559.

Cannavo A, Liccardo D, Gelzo M, Amato F, Gentile I, Pinchera B, et al. Serum galectin‐3 and aldosterone: potential biomarkers of cardiac complications in patients with COVID‐19. Minerva Endocrinol (Torino) 2022;47:270‐278.

Behnoush AH, Khalaji A, Naderi N, Ashraf H, von Haehling S. ACC/AHA/HFSA 2022 and ESC 2021 guidelines on heart failure comparison. ESC Heart Fail 2023;10:1531‐1544.

Chen M, Russo‐Neustadt A. Norepinephrine induces BDNF and activates CREB and discriminates among protein kinase C isoforms in cultured embryonic hippocampal neurons. Neurosci Med 2017;8:53‐67. doi:10.4236/nm.2017.84008

Benarroch EE. Brain‐derived neurotrophic factor: regulation, effects, and potential clinical relevance. Neurology 2015;84:1693‐1704. doi:10.1212/WNL.0000000000001507

Rasmussen P, Brassard P, Adser H, Pedersen MV, Leick L, Hart E, et al. Evidence for a release of brain‐derived neurotrophic factor from the brain during exercise. Exp Physiol 2009;94:1062‐1069. doi:10.1113/expphysiol.2009.048512

Li L, Guo H, Lai B, Liang C, Chen H, Chen Y, et al. Ablation of cardiomyocyte‐derived BDNF during development causes myocardial degeneration and heart failure in the adult mouse heart. Front Cardiovasc Med 2022;9: [eLocator: 967463]. doi:10.3389/fcvm.2022.967463

Fulgenzi G, Tomassoni‐Ardori F, Babini L, Becker J, Barrick C, Puverel S, et al. BDNF modulates heart contraction force and long‐term homeostasis through truncated TrkB.T1 receptor activation. J Cell Biol 2015;210:1003‐1012. doi:10.1083/jcb.201502100

Cannavo A, Jun S, Rengo G, Marzano F, Agrimi J, Liccardo D, et al. β3AR‐dependent brain‐derived neurotrophic factor (BDNF) generation limits chronic postischemic heart failure. Circ Res 2023;132:867‐881. doi:10.1161/CIRCRESAHA.122.321583

Cefis M, Chaney R, Quirié A, Santini C, Marie C, Garnier P, et al. Endothelial cells are an important source of BDNF in rat skeletal muscle. Sci Rep 2022;12:311. doi:10.1038/s41598‐021‐03740‐8

Pius‐Sadowska E, Machaliński B. BDNF ‐ a key player in cardiovascular system. J Mol Cell Cardiol 2017;110:54‐60. doi:10.1016/j.yjmcc.2017.07.007

Yang X, Zhang M, Xie B, Peng Z, Manning JR, Zimmerman R, et al. Myocardial brain‐derived neurotrophic factor regulates cardiac bioenergetics through the transcription factor Yin Yang 1. Cardiovasc Res 2023;119:571‐586. doi:10.1093/cvr/cvac096

Bahls M, Könemann S, Markus MRP, Wenzel K, Friedrich N, Nauck M, et al. Brain‐derived neurotrophic factor is related with adverse cardiac remodeling and high NTproBNP. Sci Rep 2019;9:15421. doi:10.1038/s41598‐019‐51776‐8

Barman HA, Sahin I, Atici A, Durmaz E, Yurtseven E, Ikitimur B, et al. Prognostic significance of brain‐derived neurotrophic factor levels in patients with heart failure and reduced left ventricular ejection fraction. Anatol J Cardiol 2019;22:309‐316. doi:10.14744/AnatolJCardiol.2019.37941

Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. Int J Surg 2021;88: [eLocator: 105906]. doi:10.1016/j.ijsu.2021.105906

Wells GA, Shea B, O'Connell D, Peterson J, Welch V, Losos M, et al. The Newcastle‐Ottawa scale (NOS) for assessing the quality of nonrandomised studies in meta‐analyses. Oxford; 2000.

Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta‐analyses. BMJ 2003;327:557‐560.

Egger M, Smith GD, Schneider M, Minder C. Bias in meta‐analysis detected by a simple, graphical test. BMJ 1997;315:629‐634.

Begg CB, Mazumdar M. Operating characteristics of a rank correlation test for publication bias. Biometrics 1994;50:1088‐1101. doi:10.2307/2533446

Fang F, Zhang X, Li B, Gan S. miR‐182‐5p combined with brain‐derived neurotrophic factor assists the diagnosis of chronic heart failure and predicts a poor prognosis. J Cardiothorac Surg 2022;17:88. doi:10.1186/s13019‐022‐01802‐0

Fukushima A, Kinugawa S, Homma T, Masaki Y, Furihata T, Yokota T, et al. Serum brain‐derived neurotropic factor level predicts adverse clinical outcomes in patients with heart failure. J Card Fail 2015;21:300‐306. doi:10.1016/j.cardfail.2015.01.003

Fukushima A, Kinugawa S, Homma T, Masaki Y, Furihata T, Yokota T, et al. Decreased serum brain‐derived neurotrophic factor levels are correlated with exercise intolerance in patients with heart failure. Int J Cardiol 2013;168:e142‐e144. doi:10.1016/j.ijcard.2013.08.073

Kadowaki S, Shishido T, Honda Y, Narumi T, Otaki Y, Kinoshita D, et al. Additive clinical value of serum brain‐derived neurotrophic factor for prediction of chronic heart failure outcome. Heart Vessels 2016;31:535‐544. doi:10.1007/s00380‐015‐0628‐6

Nakano I, Kinugawa S, Hori H, Fukushima A, Yokota T, Takada S, et al. Serum brain‐derived neurotrophic factor levels are associated with skeletal muscle function but not with muscle mass in patients with heart failure. Int Heart J 2020;61:96‐102.

Pytka MJ, Pałasz‐Borkowska A, Tarchalski JL, Nowak A, Przymuszała‐Staszak D, Schneider A, et al. The serum concentration of brain‐derived neurotrophic factor is lower in ambulatory and clinically stable patients with more advanced systolic heart failure. Pol Arch Intern Med 2022;132: doi:10.20452/pamw.16303

Shibata A, Hanatani A, Izumi Y, Kitada R, Iwata S, Yoshiyama M. Serum brain‐derived neurotrophic factor level and exercise tolerance complement each other in predicting the prognosis of patients with heart failure. Heart Vessels 2018;33:1325‐1333. doi:10.1007/s00380‐018‐1174‐9

Takashio S, Sugiyama S, Yamamuro M, Takahama H, Hayashi T, Sugano Y, et al. Significance of low plasma levels of brain‐derived neurotrophic factor in patients with heart failure. Am J Cardiol 2015;116:243‐249. doi:10.1016/j.amjcard.2015.04.018

Wu H, Cao G, Wang Y, Tian H, Du R. Increased serum CA125 and brain‐derived neurotrophic factor (BDNF) levels on acute myocardial infarction: a predictor for acute heart failure. Med Sci Monit 2019;25:913‐919. doi:10.12659/MSM.912642

Xie C, Zhan Y, Wu Y, Zhang Z, Xiang Y, Wang L, et al. Expression and clinical significance of serum sST2, BDNF, CTnI, and BUN/Cr in patients with heart failure. Altern Ther Health Med 2023;29:176‐181.

Núñez J, de la Espriella R, Miñana G, Santas E, Llácer P, Núñez E, et al. Antigen carbohydrate 125 as a biomarker in heart failure: a narrative review. Eur J Heart Fail 2021;23:1445‐1457. doi:10.1002/ejhf.2295

Friedman WJ, Ernfors P, Persson H. Transient and persistent expression of NT‐3/HDNF mRNA in the rat brain during postnatal development. J Neurosci 1991;11:1577‐1584. doi:10.1523/JNEUROSCI.11‐06‐01577.1991

Behnoush AH, Khalaji A, Khanmohammadi S, Alehossein P, Saeedian B, Shobeiri P, et al. Brain‐derived neurotrophic factor in fibromyalgia: a systematic review and meta‐analysis of its role as a potential biomarker. PLoS ONE 2023;18: [eLocator: e0296103]. doi:10.1371/journal.pone.0296103

Shobeiri P, Behnoush AH, Khalaji A, Teixeira A, Rezaei N. Peripheral levels of the brain‐derived neurotrophic factor in coronary artery disease: a systematic review and meta‐analysis. J Tehran Univ Heart Center 2023;18: doi:10.18502/jthc.v18i4.14823

Friedman WJ, Olson L, Persson H. Cells that express brain‐derived neurotrophic factor mRNA in the developing postnatal rat brain. Eur J Neurosci 1991;3:688‐697.

Khalaji A, Behnoush AH, Shobeiri P, Saeedian B, Teixeira AL, Rezaei N. Association between brain‐derived neurotrophic factor levels and obstructive sleep apnea: a systematic review and meta‐analysis. Sleep Breath 2023;27:829‐841.

Tapia‐Arancibia L, Rage F, Givalois L, Arancibia S. Physiology of BDNF: focus on hypothalamic function. Front Neuroendocrinol 2004;25:77‐107. doi:10.1016/j.yfrne.2004.04.001

Brunoni AR, Lopes M, Fregni F. A systematic review and meta‐analysis of clinical studies on major depression and BDNF levels: implications for the role of neuroplasticity in depression. Int J Neuropsychopharmacol 2008;11:1169‐1180. doi:10.1017/S1461145708009309

Suliman S, Hemmings SM, Seedat S. Brain‐derived neurotrophic factor (BDNF) protein levels in anxiety disorders: systematic review and meta‐regression analysis. Front Integr Neurosci 2013;7: [eLocator: 55]. doi:10.3389/fnint.2013.00055

Ng TKS, Ho CSH, Tam WWS, Kua EH, Ho RC. Decreased serum brain‐derived neurotrophic factor (BDNF) levels in patients with Alzheimer's disease (AD): a systematic review and meta‐analysis. Int J Mol Sci 2019;20: [eLocator: 257]. doi:10.3390/ijms20020257

Rahmani F, Saghazadeh A, Rahmani M, Teixeira AL, Rezaei N, Aghamollaii V, et al. Plasma levels of brain‐derived neurotrophic factor in patients with Parkinson disease: a systematic review and meta‐analysis. Brain Res 2019;1704:127‐136. doi:10.1016/j.brainres.2018.10.006

Feng N, Huke S, Zhu G, Tocchetti CG, Shi S, Aiba T, et al. Constitutive BDNF/TrkB signaling is required for normal cardiac contraction and relaxation. Proc Natl Acad Sci U S A 2015;112:1880‐1885. doi:10.1073/pnas.1417949112

Okada S, Yokoyama M, Toko H, Tateno K, Moriya J, Shimizu I, et al. Brain‐derived neurotrophic factor protects against cardiac dysfunction after myocardial infarction via a central nervous system‐mediated pathway. Arterioscler Thromb Vasc Biol 2012;32:1902‐1909. doi:10.1161/ATVBAHA.112.248930

Hang P, Zhao J, Cai B, Tian S, Huang W, Guo J, et al. Brain‐derived neurotrophic factor regulates TRPC3/6 channels and protects against myocardial infarction in rodents. Int J Biol Sci 2015;11:536‐545. doi:10.7150/ijbs.10754

Matsumoto J, Takada S, Kinugawa S, Furihata T, Nambu H, Kakutani N, et al. Brain‐derived neurotrophic factor improves limited exercise capacity in mice with heart failure. Circulation 2018;138:2064‐2066. doi:10.1161/CIRCULATIONAHA.118.035212

Manni L, Nikolova V, Vyagova D, Chaldakov GN, Aloe L. Reduced plasma levels of NGF and BDNF in patients with acute coronary syndromes. Int J Cardiol 2005;102:169‐171. doi:10.1016/j.ijcard.2004.10.041

Moosaie F, Mohammadi S, Saghazadeh A, Dehghani Firouzabadi F, Rezaei N. Brain‐derived neurotrophic factor in diabetes mellitus: a systematic review and meta‐analysis. PLoS ONE 2023;18: [eLocator: e0268816]. doi:10.1371/journal.pone.0268816

Dei Cas A, Khan SS, Butler J, Mentz RJ, Bonow RO, Avogaro A, et al. Impact of diabetes on epidemiology, treatment, and outcomes of patients with heart failure. JACC Heart Failure 2015;3:136‐145. doi:10.1016/j.jchf.2014.08.004

Kenny HC, Abel ED. Heart failure in type 2 diabetes mellitus. Circ Res 2019;124:121‐141. doi:10.1161/CIRCRESAHA.118.311371

Celano CM, Villegas AC, Albanese AM, Gaggin HK, Huffman JC. Depression and anxiety in heart failure: a review. Harv Rev Psychiatry 2018;26:175‐184. doi:10.1097/HRP.0000000000000162

Matthews VB, Aström MB, Chan MH, Bruce CR, Krabbe KS, Prelovsek O, et al. Brain‐derived neurotrophic factor is produced by skeletal muscle cells in response to contraction and enhances fat oxidation via activation of AMP‐activated protein kinase. Diabetologia 2009;52:1409‐1418. doi:10.1007/s00125‐009‐1364‐1

Januzzi JL Jr, Myhre PL. The challenges of NT‐proBNP testing in HFpEF: shooting arrows in the wind. JACC Heart Fail 2020;8:382‐385. doi:10.1016/j.jchf.2020.03.003

Fraty M, Velho G, Gand E, Fumeron F, Ragot S, Sosner P, et al. Prognostic value of plasma MR‐proADM vs NT‐proBNP for heart failure in people with type 2 diabetes: the SURDIAGENE prospective study. Diabetologia 2018;61:2643‐2653. doi:10.1007/s00125‐018‐4727‐7

Wu HB, Shao K, Wang YC, Wang XC, Liu HL, Xie YT, et al. Research progress of CA125 and BDNF in serum of patients with acute myocardial infarction for predicting acute heart failure. Clin Hemorheol Microcirc 2020;75:99‐106. doi:10.3233/CH‐190738