Assessment of nerve growth factor levels and clinical indicators for outcome prediction in first-episode schizophrenia

Abstract

BACKGROUND

Increasing evidence suggests that nerve growth factor (NGF) is a potential pathogenic factor for schizophrenia, as NGF levels tend to normalize following psychopathological improvement with antipsychotic treatment. However, how closely NGF changes are associated with treatment outcome remain unclear. This study investigated whether baseline and early changes in serum NGF levels could predict 8-week treatment response in patients with first-episode psychosis (FEP).

AIM

To investigate whether baseline and early changes in serum NGF levels could predict 8-week treatment response in FEP.

METHODS

This study was conducted in Yangzhou, Jiangsu Province, China. It was comprised of 78 FEP patients treated with atypical antipsychotics only and 75 sex- and age-matched healthy controls. Illness severity was measured weekly using the Positive and Negative Syndrome Scale. Serum NGF levels were measured at baseline for both groups and at 2, 4, 6, and 8 weeks for FEP patients.

RESULTS

Baseline serum NGF levels were significantly lower in the FEP group (174.21 ± 83.34 pg/mL) compared with healthy controls (219.21 ± 78.63 pg/mL; t = -3.450, P < 0.001). After antipsychotic treatment, NGF levels in the FEP group significantly increased, reaching 202.45 ± 97.61 pg/mL at week 8 (P = 0.003). In addition, multivariate logistic regression analysis identified baseline NGF [odds ratio (OR) = 1.018, 95% confidence interval (CI): 1.008-1.027], NGF at 2 weeks (OR = 1.016, 95%CI: 1.008-1.024), early NGF changes (OR = 1.088, 95%CI: 1.037-1.143), and early psychopathological improvement (OR = 7.312, 95%CI: 1.552-34.446) as significant predictors of individual treatment response in the FEP group.

CONCLUSION

Our results indicate that higher baseline NGF levels, NGF at 2 weeks after antipsychotic treatment, early NGF changes, and early psychopathological improvement can predict the response to antipsychotics in FEP patients.

Key Words: Nerve growth factor; Schizophrenia; First-episode schizophrenia; Treatment outcome; Biomarker

Core Tip: This study in first-episode schizophrenia found that baseline serum nerve growth factor (NGF) levels were significantly lower than in healthy controls and increased after 8 weeks of antipsychotic treatment. Baseline NGF, NGF at week 2, early NGF changes, and early psychopathological improvement each independently predicted treatment response. Combining all four indicators yielded superior predictive performance over any single measure, suggesting that dynamic NGF monitoring may assist early clinical decision-making.

INTRODUCTION

Schizophrenia is a serious multifactorial mental illness with potentially devastating consequences, including suicidality, cognitive impairment, unhealthy behaviors, and substantial economic burden[1]. As a major psychiatric disorder, it affects approximately 1% of the global population over a lifetime[2]. It is generally believed that the etiology of schizophrenia is associated with altered activity of neurotrophins in the brain, particularly lowered expression of brain-derived neurotrophic factor (BDNF) and nerve growth factor (NGF)[3-5]. However, compared with BDNF, NGF has been less studied in psychiatric research.

Ceci et al[6] reported that NGF plays various roles in the development of the nervous system, and other studies[7,8] showed that NGF signaling plays a vital role in nerve cell survival, proliferation, and migration. Additionally, Cuello and Do Carmo[9] reported that NGF confers neuroprotection in the hippocampus via high-affinity receptor binding, which enhances neurogenesis and inhibits apoptotic neuronal cell death. Furthermore, NGF can promote mitochondrial function[10], help with recovery of peripheral nerve injury[11], and delay decreased cognitive function in non-demented people[12,13]. Recent work also suggested that injections of NGF into the developing frontal cortex of neonatal rats can reduce social interaction, which is consistent with behaviors observed in human schizophrenia patients[14,15]. Collectively, these studies suggest that NGF may play a pivotal role in schizophrenia.

Additional studies highlighted the important role of NGF in the pathogenesis and treatment of schizophrenia[16-20]. Qin et al[3], Çakici et al[21], and Dai et al[22] reviewed recent data demonstrating that blood NGF levels were reduced in patients with schizophrenia. However, the available findings are ambiguous in terms of variations in peripheral NGF concentrations during the treatment of schizophrenia with antipsychotic medication[4,23-28]. For example, Jockers-Scherübl et al[29] reported that drug treatment increased serum NGF levels in previously medication-free patients with schizophrenia, but Ajami et al[24] reported contrary findings. In addition, Martinez-Cengotitabengoa et al[4] and Bioque et al[5] found that antipsychotic medication had no effect on NGF protein levels in patients with schizophrenia. Based on these conflicting results, whether antipsychotics can affect peripheral levels of NGF in patients with schizophrenia remains unclear.

To address this knowledge gap, we proposed two hypotheses: First, that baseline NGF levels are lower in drug-naive first-episode psychosis (FEP) patients than in healthy controls (HCs) and that antipsychotic treatment may elevate serum NGF levels; second, that baseline NGF, early changes in NGF, and early symptomatic improvement could serve as predictive biomarkers of treatment outcome. To test these hypotheses, we conducted a study to elucidate the effect and mechanism of action of antipsychotics on serum NGF levels in FEP patients, and to explore putative biomarkers of treatment response.

MATERIALS AND METHODS

Subjects

The Yangzhou Wutaishan Hospital Ethics Committee approved the study protocol. For patients who lacked decision-making capacity due to acute psychosis, written informed consent was obtained from their legal guardians. First-episode and drug-naive first-FEP subjects were recruited at Wutaishan Hospital (Yangzhou, Jiangsu Province, China) from August 2023 to January 2025. The assessment was conducted using the Structured Clinical Interview for Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition[30] and the Chinese version of the Structured Clinical Interview for DSM-IV-TR Axis I Disorders-Patient Edition, and the diagnosis of schizophrenia was confirmed according to the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition criteria. Inclusion criteria were as follows: First-FEP, age 18-60 years and antipsychotic-naïve. Exclusion criteria were as follows: Comorbid neurological or medical disorders, substance abuse (except nicotine), pregnancy/lactation, and intellectual disability. All patients underwent a complete physical evaluation and independent psychiatric assessments by two senior psychiatrists. The HC group consisted of 75 individuals recruited via local advertisements, screened by medical history, physical examination, and laboratory tests. HC exclusion criteria included physical illness, organic brain disorder, alcohol abuse, or pregnancy.

Treatment procedure

Sociodemographic and clinical data were collected through structured interviews and medical records by trained psychiatrists. Body mass index was measured, and smoking status was self-reported. After 8 weeks of follow-up, we had complete data for 78 of the 97 FEP subjects, resulting in a loss-to-follow-up rate of 19.6%. All subjects tolerated the procedures well, with no adverse events. Completers and non-completers did not differ significantly in clinical variables.

Patients were treated with different atypical antipsychotics monotherapy: 17 patients received risperidone, 5 received paliperidone, 14 received olanzapine, 17 received quetiapine, 13 received aripiprazole, 7 received amisulpride, and 5 received ziprasidone. The doses of antipsychotics were adjusted based on clinical judgment of psychiatric symptoms and were converted to chlorpromazine equivalents[31]. There were no significant differences among the seven different types of chlorpromazine equivalent dosage: F (6, 71) = 0.840, P = 0.540. The mean daily antipsychotic dose was 402.45 ± 194.32 mg.

Assessment of clinical situation

Clinical severity of schizophrenia and response to treatment were assessed using the Positive and Negative Syndrome Scale (PANSS) version with the standard 30-item manual, and results were reviewed every week by a trained rater blinded to the treatment condition. Response to treatment in the FEP group was defined as a PANSS reduction rate ≥ 50% at week 8. “Early improvement” was defined as a ≥ 20% reduction in PANSS score at week 2 relative to baseline[32].

NGF serum levels

Blood samples were obtained from all participants at baseline. In addition, FEP patients underwent follow-up sampling at weeks 2, 4, 6, and 8 of treatment. All subjects fasted for at least 12 hours prior to the blood draw, which occurred between 07:00 and 09:00. Following centrifugation of blood samples at 3500 × g for 10 minutes, the resulting serum was stored at -80°C until NGF analysis. Serum NGF levels were measured using enzyme-linked immunosorbent assay (ELISA) kits purchased from R&D Systems (Minneapolis, MN, United States) following the manufacturer’s instructions. Details of sample processing have been reported elsewhere[33]. The standards and samples were analyzed in duplicate. Measurements are expressed as pg/mL. The sensitivity of the NGF assay was 5.7 pg/mL, and inter- and intra-assay variation coefficients were 10% and 12%, respectively. Sample collection and analysis were conducted in a blinded fashion so that the study personnel analyzing the samples did not know the identity of the subject.

Statistical analysis

Descriptive statistics were obtained by calculating mean ± SD, n (%) to describe demographic characteristics. The Kolmogorov-Smirnov test was used to assess normal distribution. Normally distributed continuous variables (NGF levels and other baseline demographic and clinical characteristics shown in Table 1) of the patient and HC groups were compared using t-tests. Repeated measures analysis of variance was used to compare the differences in patients’ NGF data across five time points (weeks 0, 2, 4, 6, and 8). Pearson’s correlation test was used for all datasets with normal distribution; otherwise, Spearman’s correlation test was used.

Table 1 Demographic data and clinical characteristics of the study subjects, n (%).

	FEP group (n = 78)	HC group (n = 75)	χ2/t	P value
Gender (male/female)	40/38	42/33	0.342	0.559
Age (years)	26.03 (6.45)	26.51 (6.30)	-0.470	0.636
Education (years)	12.01 (3.18)	12.51 (3.85)	-0.870	0.382
Body mass index	22.42 (6.05)	22.46 (5.49)	-0.043	0.966
Smokers (yes/no)	28/50	33/42	1.050	0.306
Age of onset (years)	24.29 (5.27)
Duration of illness (years)	1.82 (2.14)
Family history of mental disorders (yes/no)	18/60
PANSS at baseline
Positive subscale	24.24 (8.47)
Negative subscale	21.42 (8.03)
General subscale	33.42 (3.59)
Total	79.09 (5.04)
PANSS after treatment
Positive subscale	16.09 (5.61)
Negative subscale	13.27 (4.28)
General subscale	20.08 (6.72)
Total	49.44 (11.78)

FEP: Never-medicated first-episode psychosis; HC: Healthy control; PANSS: Positive and Negative Syndrome Scale.

Logistic regression analyses were used to identify factors independently associated with treatment outcome. Treatment outcome (response vs non-response) was used as the dependent variable, and demographic data, clinical characteristics, NGF level, and the PANSS total and index scores were included as independent variables. Multicollinearity was assessed using variance inflation factor, with all variance inflation factor values below 2.0, indicating no serious multicollinearity. In addition, the receiver operating characteristic (ROC) curve was established based on NGF and other predictors in the response and non-response groups. The optimal cut-off values for continuous variables predicting treatment response were determined by maximizing the Youden index (sensitivity + specificity - 1) derived from the ROC curves.

Of the 97 FEP subjects, 19 discontinued the study (non-completers), and baseline characteristics did not differ significantly between completers and non-completers (all P > 0.05), suggesting attrition bias is unlikely. For the remaining 78 subjects with follow-up data, missing values at individual time points were handled using multiple imputation by chained equations. GraphPad Prism 9.0 was used to generate the figures. Two-tailed tests were used throughout the analyses. Statistical significance was set at P < 0.05. SPSS version 26 (IBM, Armonk, NY, United States) was used to conduct all statistical analyses.

RESULTS

Sociodemographic characteristics

Table 1 lists the characteristics of the study subjects. There were no significant differences between the patient and HC groups in terms of age, gender, education level, body mass index, or smoking status (P > 0.05).

NGF concentrations and changes with treatment

At the 2-month follow-up during antipsychotic treatment, NGF levels had significantly increased in the FEP group (Figure 1A). Baseline serum NGF levels were statistically significantly lower in the FEP group (174.21 ± 83.34 pg/mL) than in HCs (219.21 ± 78.63 pg/mL)(t = -3.450, P < 0.001) (Figure 1B). Across the five time points, repeated measures analysis of variance showed a significant main effect of time (F = 12.360, P < 0.001, Greenhouse-Geisser corrected). Post-hoc pairwise comparisons were performed with Bonferroni correction for multiple comparisons. The results indicated that only the difference between baseline NGF level (174.21 ± 83.34 pg/mL) and week 8 (202.45 ± 97.61 pg/mL) was statistically significant (P < 0.05), whereas all other pairwise comparisons did not reach statistical significance (P > 0.05).

Figure 1 Dynamic changes, intergroup differences, and stratified analysis of treatment response of serum nerve growth factor levels in first-episode psychosis patients. A: Nerve growth factor (NGF) over the five assessment time points (W0: Before antipsychotic treatment; W2: After 2 weeks; W4: After 4 weeks; W6: After 6 weeks; W8: After 8 weeks); B: Comparison of serum NGF levels between the first-episode psychosis and healthy control groups; C: Changes in baseline NGF levels after 8 weeks of medication treatment in the response group (n = 50) and the non-response group (n = 28). NGF: Nerve growth factor; FEP: First-episode psychosis patients who had never received medication; HC: Healthy controls.

Interestingly, we found that NGF at baseline and week 2 were inversely correlated with negative symptoms (r = -0.231, P = 0.042; r = -0.227, P = 0.045, respectively ), and that early NGF change was associated with negative symptoms at week 8 (r = -0.293, P = 0.010). Moreover, pre-treatment NGF level was not related to positive symptoms or to the sociodemographic data (age, sex, family history, smoking status, education, etc.) in an individual with schizophrenia (P > 0.05). In the current sample, there were no differences in NGF levels among the different types of antipsychotic drugs tested (P > 0.05).

NGF concentrations and clinical conditions

At 2 weeks after the onset of antipsychotic treatment, the PANSS scores were ≥ 20% lower in 28 cases (early improvement group) and < 20% in 50 cases (non-early improvement group). Serum NGF levels did not differ significantly between these two groups (166.40 ± 66.86 pg/mL vs 178.58 ± 89.73 pg/mL, t = -0.680; P = 0.499).

Comparison of PANSS total score at baseline vs 8 weeks of treatment revealed a significant reduction from 79.05 ± 5.04 to 49.44 ± 11.78, respectively. Fifty patients were responsive to antipsychotic treatment and 28 patients were non-responders, and a significant difference in baseline NGF was detected between these two groups (197.03 ± 78.23 pg/mL vs 108.03 ± 51.94 pg/mL, t = 4.730; P < 0.001) (Figure 1C).

We applied a binary logistic regression model to identify predictors of beneficial outcomes with specific antipsychotics in FEP patients. The results showed that baseline NGF (NGF₀) [odds ratio (OR) = 1.018, 95% confidence interval (CI): 1.008-1.027, P < 0.001], NGF at 2 weeks after antipsychotic treatment (NGF₁) (OR = 1.016, 95%CI: 1.008-1.024, P < 0.001), early NGF changes (OR = 1.088, 95%CI: 1.037-1.143, P = 0.001), and early psychopathological improvement (OR = 7.312, 95%CI: 1.552-34.446, P = 0.012) were very efficient in predicting response to antipsychotics in FEP patients (Table 2).

Table 2 Logistic regression models of treatment response at 8-week follow-up.

	B	SE	Wald	Sig	EXP (B)	95%CI for EXP (B)
						Lower	Upper
NGF0	0.018	0.005	14.024	< 0.001	1.018	1.008	1.027
NGF1	0.016	0.004	14.860	< 0.001	1.016	1.008	1.024
NGF1 - NGF0	0.085	0.025	11.662	0.001	1.088	1.037	1.143
PANSS ≥ 20% response at 2 weeks	1.990	0.791	6.331	0.012	7.312	1.552	34.446

Variance inflation factor values were all below 2.0, indicating no serious multicollinearity. CI: Confidence interval; NGF₀: Baseline nerve growth factor; NGF₁: Nerve growth factor at 2 weeks after onset of antipsychotics treatment; PANSS: Positive and Negative Syndrome Scale.

ROC curve analysis of serum NGF and other predictors of treatment outcome in schizophrenia

The aim of this study was to analyze early risk factors and protective factors of treatment outcome in patients with FEP. The ROC curve was applied to evaluate the significance of NGF₀, NGF₁, early NGF changes, and early psychopathological improvement in response to antipsychotic drug treatment. All factors were useful in predicting antipsychotic drug response. For these parameters, the area under the curve (AUC) of the ROC values were 0.830, 0.842, 0.800, and 0.674, respectively(Figure 2A-C). However, NGF at 2 weeks after treatment was most useful in terms of predictive value, followed by baseline NGF, NGF changes, and early psychopathological improvement (Table 3).

Figure 2 Receiver operating characteristic curves of serum nerve growth factor-related indicators and their combinations for predicting response to antipsychotic drug treatment. A: Receiver operating characteristic (ROC) curve of sensitivity vs specificity of baseline nerve growth factor (NGF); B: ROC curve of sensitivity vs specificity of NGF at 2 weeks after onset of antipsychotic treatment; C: ROC curve of sensitivity vs specificity of early NGF changes; D: ROC curve of sensitivity vs specificity of combinations of baseline NGF, NGF at 2 weeks after onset of antipsychotic treatment, early NGF changes, and early psychopathological improvement in response to antipsychotic drug treatment. Early nerve growth factor changes: Nerve growth factor at 2 weeks after onset of antipsychotics treatment - baseline nerve growth factor. AUC: Areas under the curve; NGF₀: Baseline nerve growth factor; NGF₁: Nerve growth factor at 2 weeks after onset of antipsychotics treatment.

Table 3 Receiver operating characteristic curve of variable sensitivity vs specificity.

Variable	AUC	95%CI	P value	Youden index	Sensitivity	Specificity	PPV	NPV
NGF0	0.830	0.728-0.906	< 0.001	0.647	89.66	75.00	91.2	71.4
NGF1	0.842	0.742-0.915	< 0.001	0.647	89.66	75.00	91.2	71.4
NGF1 - NGF0	0.800	0.694-0.882	< 0.001	0.557	70.69	85.00	93.2	50.0
Early improvement	0.674	0.559-0.776	0.021	0.348	44.83	90.00	92.9	36.0
Combined variables	0.925	0.843-0.972	< 0.001	0.716	96.55	75.00	92.8	88.2

AUC: Area under the curve; CI: Confidence interval; PPV: Positive predictive value; NPV: Negative predictive value; NGF₀: Baseline nerve growth factor; NGF₁: Nerve growth factor at 2 weeks after onset of antipsychotics treatment.

We used ROC curves and logistic regression analysis to assess the value of joint detection of NGF and other parameters in predicting treatment outcome in schizophrenia. It was more efficient than a single parameter approach (P < 0.05). Compared with the use of single parameters, which yielded sensitivities ranging from 44.83% to 89.66% and positive predictive values ranging from 91.2% to 93.2%, the combined detection of all four markers improved predictive performance, yielding a sensitivity of 96.55% and a positive predictive value of 92.80%. The AUC for the combined detection of all four markers was 0.925 (95%CI: 0.843-0.972) (Figure 2D).

DISCUSSION

Our results demonstrate that: (1) There were decreased serum NGF levels in FEP patients relative to HCs; (2) There was a change in serum NGF levels after treatment with antipsychotics; and (3) Baseline NGF, NGF at 2 weeks after antipsychotics treatment, early NGF changes, and early improvement can predict response to antipsychotic drugs in FEP patients. These findings suggest that these factors can be used as predictive markers for the response to antipsychotic drugs in patients with first-episode schizophrenia.

Baseline serum NGF level was significantly lower in FEP subjects compared with healthy subjects, consistent with previous results[3,34]. Our finding of altered peripheral blood NGF levels in FEP patients supports the neurotrophic hypothesis of schizophrenia, which suggests that abnormal neurotrophic factors play an important role in the pathogenesis of schizophrenia in these patients. Antipsychotic drugs increased NGF levels, indicating that these medications exert their effects by altering neurotrophic factor levels. However, Martinez-Cengotitabengoa et al[4], Ajami et al[24], Lee and Kim[26], among others, found no effect or even the opposite effects of antipsychotics on NGF. Possible reasons for these differences may be related to differences in patient population (first-episode vs chronic), treatment duration, detection methods, and sample size. Due to the limited sample size, differences between drugs cannot be ruled out.

Although this study identified a certain association between NGF and negative symptoms, not all studies have been consistent[19,22,35,36]. For instance, Yang et al[19] found that in 72 male patients with chronic schizophrenia, NGF-β levels were negatively correlated with the PANSS negative symptom subscale, further supporting the role of NGF in the pathological mechanism of negative symptoms. A study by Chu et al[36] involving 30 drug-naive schizophrenia patients found no significant correlation between NGF-β and any dimension of the Brief Psychiatric Rating Scale score, including negative symptoms. Similarly, Xiong et al[37] did not identify a correlation between NGF and PANSS scores. These inconsistent findings may be attributed to differences in sample characteristics, detection methods, symptom assessment tools, and statistical power.

In recent years, there has been growing interest in how neuroinflammation relates to schizophrenia. One study found that in first-FEP patients, expression of the NGF receptor TrkA increased over time. Baseline inflammation levels were linked to changes in BDNF receptor ratios, suggesting that inflammation may affect neurotrophic signaling[4]. Another review noted that schizophrenia patients have higher levels of pro-inflammatory cytokines and lower levels of neurotrophic factors like NGF and BDNF. This combination of increased inflammation and reduced neurotrophic support may contribute to the disease[38]. Lower NGF levels may weaken the brain’s ability to protect and repair itself, making it more vulnerable to inflammatory damage. Antipsychotic treatment may increase NGF levels partly through anti-inflammatory effects, helping restore balance. This idea fits with our finding that NGF levels rose and symptoms improved after treatment. Future studies should explore the link between NGF and specific inflammatory markers and whether together they can predict treatment response.

We conducted logistic regression analyses to assess the extent to which NGF levels and clinical symptoms could uniquely predict treatment outcome. We found that early response to antipsychotic drug therapy predicted subsequent clinical response in patients with schizophrenia, which is consistent with existing literature establishing early psychopathological improvement as an important protective factor of treatment outcome in schizophrenia[32,39-44]. We also found that baseline NGF, NGF at 2 weeks after antipsychotic treatment, and early NGF changes were strongly associated with treatment outcome. This is in accordance with previous reports emphasizing that higher NGF levels in patients treated with atypical antipsychotics may have implications for treatment outcome[27]. Contrary to our findings, Lee and Kim[26] reported that better response to risperidone treatment was not associated with higher levels of NGF in patients with schizophrenia. Variations in laboratory assays, study design, and limited statistical power due to small sample sizes may account for discrepancies among studies. It currently remains unclear how NGF induces antipsychotic behavior. Thus, understanding the biochemical and pathological changes in early-stage schizophrenia is critical for refining intervention strategies.

We performed ROC curve analysis to evaluate the diagnostic efficiency of baseline NGF, NGF at 2 weeks after antipsychotic treatment, early NGF changes, and early response. We detected moderate sensitivity and specificity of these factors for the discrimination of response and non-response groups. Based on the logistic model, we combined predictors and found that AUC of the ROC data were higher than that of a single parameter. Our data further revealed that combining information about NGF level and psychiatric symptoms outperformed information gained by measuring baseline NGF, NGF at 2 weeks after antipsychotic treatment, early NGF changes, or early response alone, which may provide even stronger predictors of outcome. To our knowledge, few studies have specifically explored how the prognostic relevance of NGF relates to short-term treatment outcomes in first-episode schizophrenia. Further studies of many patients with a longer follow-up would be useful to confirm our preliminary findings.

The most interesting finding of this study was that four factors independently predicted treatment response: Higher NGF levels before treatment, higher NGF levels 2 weeks after starting antipsychotics, early changes in NGF, and a good early response. Other factors showed no effect. Previous studies have suggested that NGF levels in the blood may be useful markers of treatment outcome in schizophrenia[24,45]. Therefore, these four factors could potentially help screen for which patients will respond to antipsychotic medication, even though their predictive accuracy was relatively low when tested with standard statistical methods. Adding NGF testing to traditional clinical assessment (such as the PANSS scale) gave better predictions than either method alone. However, because only 28 patients were non-responders, the results should be interpreted cautiously as they may be overfitted. If confirmed in larger studies, measuring NGF levels could give clinicians a simple and low-cost tool to support early treatment decisions and help identify patients who may need alternative treatment approaches.

This study has several main limitations. First, this was a small-sample, single-center, observational real-world study, which may be subject to selection bias and information bias. Second, the loss to follow-up rate was 19.6% during the study. Although potential bias was controlled by baseline comparison, multiple imputation and sensitivity analysis, it may still have a certain impact on the extrapolation of results. In the future, multi-center and prospective studies with optimized follow-up procedures should be conducted to further improve data integrity and conclusion reliability. Third, it is unknown whether source of peripheral NGF relate to brain NGF. Kale et al[46] reported that a significant relationship between serum NGF and cerebrospinal fluid was observed only in HCs. Thus, whether serum NGF levels are associated with the state of the brain in people with schizophrenia remains unclear. Additionally, venous blood samples to measure the levels of NGF were taken only once from participants in the HC group; the observed increase in the FEP group could reflect natural temporal variability, and future studies should include parallel longitudinal measurements in HCs to confirm a causal pharmacological effect.

CONCLUSION

Our preliminary clinical findings indicate that baseline NGF levels of patients with schizophrenia are lower than those of HCs, and antipsychotic treatment increases NGF levels. Furthermore, baseline serum NGF, NGF levels after 2 weeks of antipsychotic treatment, early changes in NGF, and early improvement in psychiatric symptoms can predict treatment outcomes in first-FEP. Combining these four indicators provides superior predictive performance compared to using any single measure alone. Therefore, testing NGF protein concentrations may assist clinicians in the early prediction of antipsychotic treatment response in patients with schizophrenia. Dynamic monitoring of NGF and its early changes could serve as a useful supplementary tool in clinical assessment.

ACKNOWLEDGEMENTS

We thank the psychiatric inpatient unit of Wutaishan Hospital of Yangzhou, China. We also express our thanks to our colleagues, who assisted in subject recruitment in this study.