In this study, our primary objective was to investigate the association between two biological markers, BDNF and CRP, among patients suffering from MMD + SA and MDD-SA. Additionally, we aimed to assess the discriminatory capabilities of BDNF and CRP for distinguishing among individuals in the MMD + SA, MDD-SA, and healthy control groups.

We observed that there were no significant differences among the three groups in terms of demographic and baseline characteristics, except for a notable difference in the occurrence of self-harm. Specifically, self-harm was more prevalent in the MMD + SA group compared to the MDD-SA group. Interestingly, the control group did not report any history of self-harm.

Regarding the biological markers, we found that CRP levels did not exhibit significant variations between the groups. However, BDNF levels showed notable differences among the groups. BDNF concentrations in MDD patients, regardless of suicide attempts, were significantly lower than those in healthy controls. Importantly, BDNF demonstrated a significant ability to predict suicide attempts when compared to healthy controls, although it did not exhibit the same predictive power when compared to the MDD group.

When evaluating the performance of BDNF in discriminating between MMD + SA and MDD-SA groups against the healthy control group, ROC curve analysis indicated excellent discriminatory power. However, BDNF could not effectively differentiate between the MMD + SA and MDD-SA groups. In contrast, ROC curves for CRP did not demonstrate significant predictive power in our analysis.

Our findings regarding the association between BDNF and suicidal behavior, however, are inconclusive compared to the literature; In line with our findings, Eisen et al. in [16] used linear regression to show that no significant association was present between suicide attempt and BDNF level. Also, Eisen et al. in 2015 published a meta-analysis stating no significant correlation between BDNF levels and suicide attempts [17]. Moreira et al. [18] also indicated that BDNF levels were significantly lower in MDD patients with or without suicide attempts compared to healthy controls but no significant difference was found between depressed patients regardless of their suicide attempt history. Accordingly, it could be hypothesized that BDNF could be an indicator of depression not related to suicidal actions, and suicide attempts which not go on to complete suicide wouldnot have lower BDNF than depressed patients without suicide attempts.

Contrary to our results, Khan et al. [14] reported that even suicidal ideation in depressed patients is significantly correlated with lower BDNF levels compared to MDD-SA patients. Regression analysis also confirmed that no confounding factor influenced this significant relationship between BDNF and suicidal thoughts. Ai et al. [8] also showed that not only BDNF was significantly lower in MDD patients compared to controls, but also it was significantly lower in MDD + SA patients compared to MDD-SA patients. In that study, a significant correlation between BDNF concentration and suicidal thoughts or suicide attempts was found but BDNF level was not correlated with depression severity. Kudinova et al. [19] also emphasized the lower levels of plasma BDNF in cases with suicide attempt history compared to patients without any suicide attempt history. Both Ai et al. and Kudinova evaluated a history of suicide attempts not a recent history of suicide attempts as our participants. For instance, no participant in the Kudinova article reported a suicide attempt within the last year.

The discriminatory properties of BDNF in distinguishing between MDD + SA and MDD-SA groups have been explored in various studies, yielding different results. In our study, adjusted ROC curves based on regression models for BDNF exhibited excellent discriminatory ability for distinguishing the healthy control group from both the MMD + SA group (AUC = 0.754; 95% CI 0.877–0.632; sensitivity = 73.3%; specificity = 70%; cutoff = 18.43) and the MDD-SA group (AUC = 0.796; 95% CI 0.909–0.683; sensitivity = 73.3%; specificity = 70%; cutoff = 18.51). However, these adjusted ROC curves could not effectively differentiate between individuals with suicide attempts and those without (p = 0.313).In a study by Khan et al. [14], the most suitable cutoff point for BDNF levels between individuals with suicidal depression and those without was determined to be 444.58 pg/ml, with a sensitivity of 68.7% and specificity of 78.1%. However, no significant difference in BDNF levels was observed between the depressive control and normal control groups (p = 0.996).Lee et al. [20] reported that BDNF levels between MDD and normal control groups exhibited good discriminatory power using ROC curve analysis (AUC = 0.774; 95% CI 0.663 to 0.884; sensitivity = 78.3%; specificity = 81.3%; cutoff = 684.75). Moreover, suicidal MDD was significantly correlated with low BDNF levels, with an odds ratio of 33.123.

Regarding the discriminatory properties of BDNF in distinguishing between MDD and healthy control (HC) groups, Chiou and Huang [21] found that BDNF levels had poor discriminatory efficacy for depressed patients and HCs (AUC = 0.562, 95% CI 0.516 to 0.607). The optimal cutoff point for BDNF levels was 6.02 ng/ml, with a sensitivity of 72.8% and specificity of 41.8%. However, BDNF levels demonstrated moderate diagnostic power in the male subgroup (AUC = 0.652, sensitivity = 81.1%, and specificity = 48.5% at the BDNF level of 5.11 ng/ml) but not in the female subgroup (AUC = 0.536).In the study by Shahyad et al. [22], BDNF levels exhibited suitable discriminatory efficacy for distinguishing between depressed patients and HCs (AUC = 0.823; 95% CI 0.935–0.711; sensitivity = 80%; specificity = 83%; cutoff = 1206).

These findings suggest that the discriminatory properties of BDNF can vary across different studies and populations, highlighting the complexity of its role in distinguishing between various psychiatric conditions.

Altogether, our study in line with numerous previous reports [16,17,18], recognizes BDNF as a factor involved in the development of MDD without any association with suicide attempt while other studies, correlate BDNF levels with suicide attempt itself [8, 14, 19]. Our findings confirm the role of BDNF in the pathophysiology of depression but the correlation of BDNF and suicide is yet to be elucidated. It seems that these inconsistencies between studies originate from differences in study designs as depression severity; sample size, depression duration, and other parameters are probably different between studies and lead to these discrepancies. Importantly, ethnicity diversities should be considered and more evaluations ofdifferent regions are required for more reliable conclusions.

CRP levels were slightly different between the three groups of our study but these differences did not reach statistical significance level.CRP was not capable of predicting suicide or even depression noticeably. Our findings in this regard are not consistent with the majority of available studies. For instance, Goalkp et al. [23] have reported that CRP concentrations were significantly higher in the suicide group compared to controls. Kumar et al.[24] also confirmed this finding. Mohamed et al. [25] showed that even the presence of suicidal thoughts is significantly associated with CRP elevation. Courtlet et al.[26] also revealed that not only CRP is elevated in MDD + SA patiant but this relationship acts in a dose–response manner and increases in CRP levels are associated with a higher risk of suicide. Gibbs et al. [27] have also confirmed this linear relationship in which the level of CRP is positively correlated with lifelong suicide attempt frequency but it was not associated with suicidal thoughts.

As it can be concluded from the number of studies in recent years, abnormal inflammatory response has been a matter of focus in suicide research field. As a systemic inflammatory marker, association of CRP and MD + SA group can be somewhat result of a process in which increased CRP concentration can enhance blood–brain barrier permeability followed by entrance of CRP into the central nervous system and posing direct and indirect influences on CNS. In addition, prior research has shown that CRP can induce reactive activation of microglia and astrocytes and also rapid proliferation of glial cells which ultimately leads to neurons’ injury [28] These underlying processes may explain the association of CRP and suicide mentioned in other studies. Relationship of CRP and suicide was not observed in our study which could be due to small sample size and different study design. CRP can be confounded by various factors such as obesity, smoking, low vitamin D levels, low physical activity, inappropriate diet, allergy, stress, sleep disturbances and subclinical infections [29] which were not completely controlled in our study.

In the initial crude analysis, sex did not serve as a predictor. However, in the final model, being female was associated with higher odds of belonging to the MDD + SA and MD-SA group groups. This observation can be potentially explained by the higher prevalence of depression in women. Women tend to make suicide attempts more frequently than men, although they are more inclined towards attempting suicide rather than completing it. In contrast, men are more likely to carry out suicide and often employ more violent methods. Consequently, women can be characterized as the “attempters” and “survivors” of suicide attempts [30].

In this study, an individual's height was found to elevate the odds of being part of both the MDD and MDD + SA groups. When it comes to the connections between depression and height, the existing body of research presents a limited and conflicting set of findings. Some researchers propose that height, particularly shorter stature, may act as a causal risk factor for depression [31], while others hold a differing viewpoint [32, 33]. AlsoThere was inverse association between height and suicide riskin some studies [34]. Height serves as an indicator of health status, susceptibility to specific illnesses, and overall quality of life. The prevalent inclination to link taller stature with notions of physical attractiveness, authority, leadership, enhanced cognitive abilities, and achievements in both social and professional domains fosters the belief that, in the context of human growth, akin to economic growth, larger is deemed superior [35]. Individuals of shorter stature are more prone to occupy lower social class positions in adulthood, irrespective of their childhood social class. This lower social class status is linked to an increased risk of suicide [36]. These insights could provide a rationale for the observed connection between height and suicide attempts or MDD as highlighted in certain research studies.Regarding the inconsistency in research findings, it is plausible to suggest that there might be undisclosed moderating variables influencing the relationship between height and depression.

In the initial crude analysis, marital status (being single) did not emerge as a predictive factor. However, in the final model, being single was associated with a decreased likelihood of being in the MDD group. These outcomes do not align with previous findings in this context [37, 38]. In considering a potential explanation for this discovery, it’s worth noting that various variables such as the quality of relationships, cultural factors, and socioeconomic status could potentially act as moderators in the relationship between MDD and marital status.

In this study, a noteworthy difference was observed in the history of self-harm among the various groups.Importantly, none of the participants in the control group reported any history of self-harm. while self-harm is not limited exclusively to individuals with depression, it is notably more prevalent among individuals who experience depression or those who have attempted suicide. [39]; When considering potential explanations for this finding, it’s conceivable that the choice of sample source, whether drawn from the general population or admitted psychiatric patients, could have influenced the study’s outcomes. Additionally, the unique religious beliefs and cultural context in Iran may play a role. In Iranian society, self-harm is both forbidden and stigmatized, which may lead to underreporting of such behaviors among the general population [40].

To our knowledge, this study is the first to simultaneously assess BDNF and hs-CRP levels in Iranian patients with MDD + SA and MD-SA group, all of whom were drug-free. We excluded individuals who had previously used mood stabilizers or antidepressants due to their potential impact on BDNF and hs-CRP levels. However, there are limitations: our study design couldn’t establish causal relationships between the parameters, and we didn’t consider factors like sleep problems, socio-demographics (rural vs. urban living), seasonality, or illness duration and severity, all known to affect BDNF expression [22].Due to the illegality of alcohol consumption in Iran and potential inaccuracies in participants’ self-reporting, alcohol usage was not considered in our assessment. The patients were not monitored for the occurrence of other mental episodes, primarily because suicide patients, who were hospitalized, were less inclined to participate in the research. Motivating participation in questionnaire completion posed a challenge, and we took care not to seek consent from heavily medicated or emotionally distressed patients incapable of providing informed consent. In such instances, our study personnel contacted the hospital later when the patient was stable. Furthermore, we did not analyze post-mortem blood samples from individuals who completed suicide due to hospital regulations. Consequently, the generalization of our findings to all suicide cases should be done cautiously. Lastly, we measured serum BDNF levels, recognizing that BDNF can traverse the blood–brain barrier, and there is a suggested positive correlation between serum and cortical BDNF levels.