Neurotrophins are involved in the development, maintenance, and plasticity of neurons (1), as well as in the development of the heart and blood vessels (2–4). Among the neurotrophins, brain-derived neurotrophic factor (BDNF), a crucial mediator of homeostasis and pathogenesis in the cardiovascular system, has garnered considerable attention (5).

Previous studies have reported a relationship between altered BDNF signaling pathways and CVD. Patients with acute coronary syndrome (ACS) showed reduced circulating levels of BDNF compared with control subjects in a cross-sectional study (6). Higher circulating BDNF levels were longitudinally related to a lower risk of CVD and mortality in a general population cohort (7). As in the general population, lower circulating BDNF levels negatively affected clinical outcomes in patients with angina pectoris (8) or heart failure (9).

A decrease in CpG methylation within the regulatory region of the Bdnf gene has been linked with increased production of BDNF in neurons (10). Given the findings of prior clinical studies exploring the link between BDNF and CVD (6–9), it is likely that BDNF hypermethylation is linked to the development of CVD and a poor prognosis. In our previous study, a higher BDNF methylation level was associated with composite major adverse cardiovascular events (MACEs) in ACS patients (11). Since sBDNF and BDNF methylation levels have been linked to cardiovascular outcomes in previous research (8, 9, 11), and there is a potential link between sBDNF and BDNF methylation levels (10), it is anticipated that these two may interact to affect cardiovascular outcomes. However, no research has been done on this matter.

Using information from a prospective study of ACS patients in Korea, we examined the modifying effect of the serum BDNF (sBDNF) level on the association between BDNF methylation and long-term cardiovascular outcomes.

In total 969 (84.1%) of the 1,152 patients examined at baseline gave their consent to offer blood samples (Supplementary Figure S1). Between those who agreed to submit blood samples and those who declined, the baseline values were not substantially different. All participants were followed to assess cardiovascular outcomes for 5–12 years, until 2017 or death [median; mean (standard deviation) duration of follow-up = 8.4; 8.7 (1.5) years].

The levels of sBDNF were 17.6 (9.4) ng/mL for the median (interquartile range), and 17.8 (7.0) ng/mL for the mean (standard deviation). In the 969 study participants, there was no correlation between the baseline sBDNF level and the average BDNF methylation level (r² = 0.001, P = 0.969). Since the BDNF expression is regulated by a genetic polymorphism entailing substitution of valine by methionine at codon 66 (Val66Met) in the pro-BDNF molecule (19), we compared the sBDNF level and BDNF methylation according to the presence of BDNF Val66Met polymorphism (Supplementary Table S1). sBDNF level was lower in Val/Met and Met/Met genotype compared to Val/Val genotype. In addition, BDNF methylation was higher in Val/Met genotype compared to Val/Val genotype. Baseline characteristics according to sBDNF level are summarized in Supplementary Table S2. A low sBDNF level was significantly associated with older age, higher frequency of Killip class > 1, and higher CK-MB level. Referencing the system that produced the data and any potential collinearity between the variables (18), 11 parameters (age, sex, Beck Depression Inventory score, depression comorbidity and treatment, previous history of ACS, diabetes, hypertension, hypercholesterolemia, obesity, smoking, and Killip class) were included as covariates in the adjusted analysis [see Supplementary Figure S2 for a directed acylic graph (20–22)].

In the low sBDNF group (n = 484), the primary outcome (composite MACEs) occurred in 212 participants (43.8%); of the secondary outcomes, all-cause mortality occurred in 104 (21.5%) participants, cardiac death in 59 (12.2%), MI in 55 (11.4), and PCI in 76 (15.7%). In the high sBDNF group (n = 485), the primary outcome occurred in 171 participants (35.3%), and the secondary outcome all-cause mortality occurred in 74 (15.3%), cardiac death in 39 (8.0%), MI in 46 (9.5%), and PCI in 63 (13.0%). Figure 1 illustrates the cumulative risk of composite MACEs in subjects with lower vs. higher average BDNF methylation levels according to the sBDNF level. In the low sBDNF group, a significant difference was observed: the composite MACE incidence was 32.8% (76/232) in those with lower and 54.0% (136/252) in those with higher methylation levels [log-rank P-value < 0.001] (Figure 1A). In addition, significant differences in those with lower vs. higher methylation levels were observed in the incidence of all-cause mortality [15.1% (35/232) vs. 27.4% (69/252), log-rank P-value = 0.002] and PCI [12.1% (28/232) vs. 19.0% (48/252), log-rank P-value = 0.018] (Supplementary Figure S4). Effects of a higher BDNF methylation level on composite MACEs and all-cause mortality were seen in the adjusted analysis (Table 1). However, in the high sBDNF group, significant differences in the primary or secondary outcomes were not observed (Figure 1B, Table 1, and Supplementary Figure S5). In the main analysis, the interaction effect between the sBDNF and average BDNF methylation levels on composite MACEs or all-cause mortality was significant after adjusting for covariates (Table 1). When the average BDNF methylation level was treated as a continuous variable, the interaction effect between the sBDNF and average BDNF methylation level on composite MACEs was also significant (Supplementary Table S3). Similar results were observed when both sBDNF and average BDNF methylation levels were treated as continuous variables (Supplementary Table S4). Considering the competing risk situation, we re-analyzed after excluding patients with non-cardiac death before the onset of MI or PCI (Supplementary Table S5) and excluding patients with death before the onset of MI or PCI (Supplementary Table S6), respectively. Generally, similar results were observed to the main analysis. All model assumptions were all met (Schoenfeld P-values > 0.3). The sBDNF level had no effect on the incidence of the primary or secondary outcomes in the adjusted analysis (Supplementary Table S7).

In this study, we found a modifying effect of the sBDNF level on the association between BDNF methylation and long-term cardiovascular outcomes using data from a prospective study of Korean ACS patients. A higher BDNF methylation level in patients with low sBDNF levels was a significant predictor of poor long-term cardiovascular outcomes such as composite MACEs, all-cause mortality, and PCI. These results remained reliable after accounting for relevant covariates. However, a higher BDNF methylation level had no influence on long-term cardiovascular outcomes in patients with high sBDNF levels.

The methylation status of the BDNF genomic region that we evaluated corresponds to an equivalent region in rat Bdnf, which was variably methylated and related to Bdnf mRNA expression (23, 24). The regulatory region of the Bdnf gene's CpG methylation has been shown to be lessened in neurons with increased BDNF synthesis (10). In our previous study, a higher average BDNF methylation status was associated with an increased incidence of composite MACEs in ACS patients (11). Because BDNF methylation status is linked to neuronal BDNF production, baseline circulating BDNF level may have an impact on these effects. In this study, higher average BDNF methylation had a negative impact on long-term cardiac outcomes only in the low sBDNF group. The interaction effect of sBDNF level and average BDNF methylation on composite MACE was sustained even when both variables were analyzed as continuous variables. These results might be explained by the synergistic effect of both unfavorable exposures (higher BDNF methylation and low sBDNF level). This theory is supported by the fact that higher BDNF methylation had less of a negative impact in the high sBDNF group. Furthermore, as this study found no correlation between sBDNF level and BDNF methylation, we surmise that the adverse effects of both exposures, which are independent of one another, work in concert.

As mentioned in the Introduction, BDNF stimulates a variety of circulatory system cell types and regulates the growth and dynamics of the cardiovascular microcirculation (25). Based on those results, the relationship between circulating BDNF level and CVD has been investigated. ACS patients showed lower sBDNF levels compared with control subjects in a cross-sectional investigation (6), and a lower sBDNF level was longitudinally associated with CVD incidence and mortality in a general population cohort (7). Furthermore, decreased circulating BDNF levels were linked to a worse clinical outcome in a study of individuals with CVDs such as angina pectoris (8) or heart failure (9). In our study of ACS patients, the sBDNF level, on the other hand, had no effect on long-term cardiovascular outcomes. Differences in the type of underlying disorder could explain why our findings differed from those of previous studies. In our investigation, sBDNF levels were measured within 2 weeks of ACS onset. Because ACS is associated with acute psychosocial stress (26), and the sBDNF level rises in response to acute psychosocial stress (27), the sBDNF level may not have had an effect on long-term cardiovascular outcomes in ACS patients due to overall elevated sBDNF levels. However, because we did not assess the concentration of sBDNF in relation to numerous CVDs in our study, more research is needed to corroborate this.

In interpreting our findings, it is important to take into account a number of study limitations. First, despite this location's prior evaluation (23, 28), the methylation status of just one CpG island in BDNF was assessed. Second, only 84% of the baseline sample could undergo methylation analysis due to attrition during the recruitment procedure. However, there were no differences in the baseline demographic and clinical traits of patients with and without access to this information. Third, while the study hypotheses were founded on prior research, the results lacked mechanistic support, demanding additional study. Fourth, although chronic kidney disease is known to be linked to BDNF levels (29) and MACEs (30), this information was not accessible for the study participants. Fifth, it was unclear which tissue is the major source of BDNF detected in the serum. Finally, the study was restricted to one institution, which limits its generalizability but is a benefit because it assures consistency in patient assessment and care.

This study has a number of strengths. It is the first prospective study to look at how sBDNF and average BDNF methylation levels interact with regard to long-term cardiovascular outcomes in ACS patients. All eligible patients who had recently had an ACS episode were enrolled as participants at baseline, reducing the likelihood of error caused by varying the testing times and increasing sample homogeneity. All psychiatric and cardiovascular assessments were conducted using well-validated measures. In addition, various covariates were considered in the analyses.

A higher average BDNF methylation status during the acute phase of ACS was associated with worse long-term cardiovascular outcomes only in patients with low sBDNF levels, not in those with high sBDNF levels, regardless of the relevant covariates. These results suggest that combining the BDNF methylation and sBDNF levels may help identifying ACS patients who are likely to have unfavorable clinical outcomes. From a therapeutic perspective, patients with BDNF hypermethylation and low sBDNF levels require special attention. However, future prospective studies are needed to ascertain whether giving these individuals extra attention results in an improved prognosis for ACS patients.

The original contributions presented in the study are included in the article/Supplementary material, further inquiries can be directed to the corresponding author.

The studies involving human participants were reviewed and approved by Institutional Review Board of Chonnam National University Hospital. The patients/participants provided their written informed consent to participate in this study.

WC and J-MK: conceptualization, data curation, formal analysis, and writing. H-JK: data curation and methodology. J-WK: formal analysis and methodology. HK, H-CK, S-WK, J-CK, YA, and MJ: data curation, validation, and project administration. All authors contributed to the article and approved the submitted version.