Enhanced niacin skin flushing response in perinatal compared to non-perinatal depression

Abstract

BACKGROUND

Perinatal depression (PND) accounts for a substantial proportion of maternal morbidity, yet its diagnosis still depends entirely on clinical symptoms, with no biological measure available in routine practice. How PND differs from non-perinatal depressive disorder (DD) at a biological level remains unclear, and no objective marker exists to differentiate them clinically. The niacin skin flushing response (NSFR) is a functional biological marker linked to lipid metabolism and inflammatory pathways. This study examined whether NSFR can differentiate PND from DD and explored associated clinical and dietary factors.

AIM

To clarify whether NSFR can be used as a specific biomarker to discriminate between PND and DD.

METHODS

One hundred and twenty-three treatment-naive first-episode patients were enrolled from Shanghai Changning Maternity and Infant Health Hospital: 67 with PND and 56 with DD. Clinical severity was assessed with the Hamilton Anxiety Rating Scale (HAMA), Hamilton Depression Rating Scale and Edinburgh Postnatal Depression Scale (EPDS). Each participant completed an NSFR test and a food frequency questionnaire. Between-group comparisons and correlation analyses were conducted.

RESULTS

Compared with the DD group, the PND group showed significantly elevated NSFR scores (P < 0.001), with more pronounced NSFR elevation in severe PND cases (P = 0.035). Within the PND group, postpartum depression severity (EPDS) was inversely associated with NSFR intensity (r = -0.539, P = 0.014), symptom severity in patients with DD did not show any meaningful association with NSFR. Depressive and anxiety symptoms correlated positively with gestational age and the interval between delivery and consultation (all P < 0.05), and were more severe postpartum than prenatally (P < 0.05). Higher intake of dark and light vegetables and eggs was associated with milder postpartum depression (all P < 0.05), whereas greater poultry intake correlated with more pronounced depressive symptoms (r = 0.358, P = 0.008) and anxiety (r = 0.341, P = 0.011).

CONCLUSION

A hyperactive NSFR distinguishes PND from DD, with specific risk and dietary factors identified. NSFR represents a promising biomarker for differentiating these conditions.

Key Words: Perinatal depression; Niacin skin flushing response; Non-perinatal depression; Nutrition; Biomarker

Core Tip: Perinatal depression (PND) is associated with a hyperactive niacin skin flushing response (NSFR) compared to non-perinatal depressive disorder (DD). NSFR may be useful as a biomarker for distinguishing PND from DD. Advanced maternal age, later pregnancy and postpartum stages, delayed help-seeking, and high poultry intake are identified as risk factors for PND, whereas vegetable and egg intake appear to have a protective effect against postpartum depressive symptoms.

INTRODUCTION

Depression is diagnosed solely on the basis of clinical symptoms. At present, there are no universally accepted and validated biomarkers available for its auxiliary diagnosis and the selection of therapeutic regimens[1]. Perinatal depression (PND) develops during pregnancy or within one year postpartum, posing severe adverse impacts on the health of mothers and their offspring[2]. PND is not only closely correlated with obstetric complications, adverse delivery outcomes, and developmental impairment in children and adolescents[3-5], but also aggravates family conflicts and elevates the risk of depression in fathers[6,7]. Although PND is characterized by hormonal fluctuations, immune dysfunction and psychosocial stress that are absent in non-PND[8], its current diagnostic criteria remain identical to those for major depression, both relying exclusively on clinical manifestations. Accordingly, exploring specific biomarkers for PND is highly valuable for improving the early detection process, diagnosis and clinical management of this disorder.

Niacin skin flushing response (NSFR) is an “in vivo functional phenotype” mediated by the activation of G-protein coupled receptor 109A (GPR109A), which subsequently triggers prostaglandin release. It is currently recognized as a potential biomarker for psychiatric disorders[9,10]. Most previous studies have found that patients with depression, schizophrenia and bipolar disorder exhibit attenuated or delayed NSFR, which is attributed to neuroimmune dysregulation and abnormal metabolic function[11,12]. Dramatic fluctuations in endocrine hormones occur during the perinatal period, thereby altering the activity of G-protein coupled receptors[13]. Therefore, this study hypothesizes that patients suffering from PND exhibit a modified NSFR, and its changing characteristics differ from those of non-PND [depressive disorder (DD)].

Nutritional factors can affect emotional states through pathways such as neurotransmitter transmission, neuroinflammation, and mitochondrial function[14], yet in-depth research on their specific role in PND remains lacking at present. This study comparatively analyzed the demographic characteristics, clinical manifestations, dietary patterns, and niacin skin flush reaction phenotypes of patients with PND and non-PND, aiming to explore the differential characteristics between the two types of disorders at the biological level.

MATERIALS AND METHODS

Participants

Eligible participants were enrolled between July 2023 and December 2024.

Inclusion criteria: (1) Female participants aged 18-60 years; (2) Meeting the diagnostic criteria for a depressive episode as defined by the International Classification of Diseases, 10^th Revision; (3) First depressive episode, no use of antidepressants within the latest two months, and no history of receiving physical treatments including electroconvulsive therapy (ECT), modified ECT, and repetitive transcranial magnetic stimulation; and (4) With intact cognitive function, capable of understanding the research protocol and completing all assessments independently.

All enrolled individuals willingly signed written consent forms following a full explanation of the study details, either by themselves or their legal guardians.

Exclusion criteria: (1) Complicated with severe physical diseases or other psychiatric disorders; (2) Having a history of alcohol or substance abuse and dependence; (3) Previous history of manic episodes, hypomanic episodes or recurrent depressive episodes; (4) Application of non-steroidal anti-inflammatory drugs, glucocorticoids, vitamins, immunosuppressants and other related medications within the past two weeks; (5) Having a severe allergic history, or currently diagnosed with skin diseases and immune system diseases; and (6) Unable to independently cooperate with the procedures required as part of this work.

The study subjects were grouped into the PND group (patients with depression onset during pregnancy or within one year after delivery) and the non-PND group.

Sample size calculation

We calculated the sample size on the basis of prior studies indicating that PND is biologically distinct from other depressive conditions[15,16]. G Power software was used for sample size estimation. With an effect size of 0.5, using a two-tailed α of 0.05 and 80% statistical power, we estimated that at least 53 participants were needed per group. In this study, 65 participants were preset to be recruited in each group to reserve allowance for case dropout. During the recruitment period, the number of patients meeting the inclusion criteria for PND exceeded expectations. Therefore, all eligible participants were enrolled, and a final total of 67 cases met the inclusion criteria for the PND group. The non-PND group was restricted by a fixed recruitment period and stricter exclusion criteria (e.g., previous use of antidepressants, combined physical diseases, etc.), with 56 cases enrolled ultimately, which was still higher than the minimum sample size requirement of 53 cases per group. Post-hoc power analysis was carried out using the actual sample size, and this gave a power of 0.85, which surpassed the commonly accepted level of 0.80. It indicated that the study cohort size in the present study had sufficient statistical testing power.

General data collection

Demographic and clinical variables were collected, including hospital identification number (inpatient or outpatient), sex, age, body mass index, pregnancy status, gestational age or postpartum duration, age at first onset, diagnosis and treatment history, duration of current episode, smoking and alcohol consumption status, and educational attainment.

Assessment of dietary intake

Nutrient intake across the month prior was examined using a tested and confirmed food frequency questionnaire[17], covering the frequency and quantity of consumption across 24 food categories. These included whole and mixed grains, eggs, meat, aquatic food from rivers and oceans, legume-derived items, and vegetables of varying shades (dark and light), dairy products, fruits, sweets, drinks, and alcohol.

Assessment of emotional symptoms

To evaluate the severity of depressive, anxiety, and somatic symptoms, the following scales were administered.

The 24-item Hamilton Depression Rating Scale: The 24-item Hamilton Depression Rating Scale[18] contains 24 items. Most items adopt a “5-point scoring method” (0 = absent, 1 = mild, 2 = moderate, 3 = severe, 4 = extremely severe), while a small number of items adopt a “3-point scoring method” (0 = absent, 1 = mild, 2 = severe). The total score of the scale ranges from 0 to 72 points, with increased scores corresponding to more pronounced depressive manifestations.

The Hamilton Anxiety Rating Scale: The cutoff criteria for classifying the severity of depressive symptoms according to the total score were as follows: A combined score of less than 8 points indicated absence of depressive manifestations; 8-20 points suggested possible depression; 21-35 points confirmed definite depression; and more than 35 points was defined as severe depression. And the Hamilton Anxiety Rating Scale[19] consists of 14 items, every item evaluated on a 0-4 scale (5 points in total), with a cumulative score range of 0-56 points. Totally score of ≥ 14 points to the existence of anxiety symptoms that are clinically important.

Edinburgh Postnatal Depression Scale: The Edinburgh Postnatal Depression Scale (EPDS)[20] is a “10-item self-rating scale” designed for perinatal populations. Each item of the scale is rated using a 0-3, 4-point scoring system, with an overall score array of 0-30 points. Items differ in scoring direction: For most items, a score of 0 indicates the absence of relevant symptoms (e.g., “able to cope with daily affairs as usual”); while for “reverse-scored items” (items 1, 2 and 10), a score of 0 corresponds to the most severe symptom level. A higher total score of the scale suggests more severe postpartum depressive symptoms. The cutoff criteria are as follows: In total score ≥ 10 points suggests a potential vulnerability to depression; a total score ≥ 13 points strongly suggests a major depressive episode.

Generalized Anxiety Disorder-7: A 7-item self-rating scale (Generalized Anxiety Disorder-7)[21] applied for measuring anxiety symptoms for the past 14 days. Each item is scored from 0 to 3 points, with the full score spanning from 0 to 21 points. Cutoff scores of 5, 10, and 15 points correspond to low, medium, and high degrees of anxiety, correspondingly.

Patient Health Questionnaire-9: A 9item scale [Patient Health Questionnaire-9 (PHQ-9)][22] completed by the participant applied to determine depressive symptoms in the preceding two weeks. A 0-3 scale is used to rate each item, and the total score ranges from 0 to 27 points. The grading criteria for symptom severity are as follows: 0-4 points for minimal severity, 5-9 points for mild severity, 10-14 points for moderate severity, 15-19 points for moderately severe severity, and 20-27 points for severe severity.

Somatic Symptom Scale-8: It is an 8-item self-reported scale [Somatic Symptom Scale-8 (SSS-8)][23,24] created to gauge somatic symptoms over the “past week”. Every item receives a score from 0 to 4, and the summed total can range from 0 to 32.

Stratification criteria for symptom severity are defined as follows: Scores of 0-3 points indicate no or minimal somatic symptoms; 4-7 points reflect a mild level of symptomatology; 8-11 points correspond to moderate symptom severity; 12-15 points indicate moderately severe symptoms; and 16-32 points represent a severe burden of somatic symptoms. According to the total SSS-8 score and clinical manifestations, the study participants were divided into mild, moderate, and severe subgroups.

NSFR test

A integrated niacin skin reaction detection device (TY-AraSnap, TY-AraPatch; Shanghai Tianyin Biotechnology Co., Ltd.) was adopted to quantitatively measure the NSFR. The standardized operational procedure was detailed as follows: (1) A label marked with the personal serial number was attached to the inner aspect of the subject’s left forearm firstly, followed by application of patches preloaded with methyl nicotinate (AMN) aqueous solution at 6 gradient concentrations, and a 1-minute timer was started immediately afterwards; (2) After 1 minute, the central portion of each patch was removed while the outer border remained in place. Automatic continuous imaging was activated, and a 5-minute countdown began; images of the forearm were captured every 10 seconds, yielding 30 images per participant; (3) At the end of 5 minutes, participants withdrew their arm and patch borders and labels were removed; and (4) Flushing intensity was quantified by comparing each post-application image to the baseline (first image). Each participant generated 174 raw flushing scores (6 AMN concentrations × 29 post-application time points); total NSFR intensity was the sum of these scores.

Statistical analysis

All analyses were conducted employing IBM SPSS Statistics version 23.0. We present continuous variables as mean ± SD. The independent-samples t-test was employed for normally distributed data, while the Mann-Whitney U test was used for data not conforming to a normal distribution, to compare differences between groups. For categorical variables, comparisons were made using Pearson’s χ² test or Fisher's exact test when expected cell counts fell below 5. Variable associations were examined with Pearson correlation continuous data following a Gaussian distribution and Spearman’s rank correlation coefficients for non-Gaussian continuous data or ordinal measures. Between-group differences in NSFR were evaluated using multivariable logistic regression adjusting for demographic and clinical covariates that differed significantly at baseline. We considered a two-tailed P < 0.05 as indicating statistical significance.

RESULTS

Demographic and clinical characteristics of participants

Totally of 123 patients were sourced from the psychological unit of Shanghai Changning Maternity and Infant Health Hospital, including 67 diagnosed with PND and 56 with DD. Marked intergroup differences were detected in several baseline characteristics. The PND group had a higher mean age than the DD group (30.7 ± 3.9 years vs 29.4 ± 8.0 years, P = 0.036) and exhibited a greater body mass index (21.8 ± 3.0 kg/m² vs 20.5 ± 2.8 kg/m², P = 0.015). The duration of illness was significantly less in the PND group compared with the DD group (13.9 ± 30.5 months vs 30.3 ± 39.0 months, P < 0.001). Depression severity differed substantially between groups: The PND group had 27 mild, 21 moderate, and 19 severe cases, with 59 patients carrying comorbid anxiety; the DD group had 9 mild, 20 moderate, and 27 severe cases, all 56 with anxiety symptoms (Table 1).

Table 1 Demographics and clinical characteristics, mean ± SD.

		PND group (n = 67)	DD group (n = 56)	P value
Age (year)		30.7 ± 3.9	29.4 ± 8.0	0.036a
BMI (kg/m2)		21.8 ± 3.0	20.5 ± 2.8	0.015a
Depression severity (n)	Mild	27	9
	Moderate	21	20
	Severe	19	27
Anxiety severity (n)		59	56
Disease duration (year)		13.9 ± 30.5	30.3 ± 39.0	< 0.0001a
Smoke (%)		9.0	14.3	0.354
Alcohol consumption (%)		10.4	17.9	0.236

^aP < 0.05.

PND: Perinatal depression; DD: Depressive disorder; BMI: Body mass index.

Intergroup comparison of NSFR

Mean NSFR intensity was 1580.3 ± 342.7 in the PND group vs 1256.9 ± 298.4 in the DD group, a difference that remained significant after multivariable adjustment (P < 0.001, Figure 1).

Figure 1 Comparison of niacin skin flushing response intensity between the perinatal depression and depressive disorder groups. Between-group differences were analyzed using multivariable logistic regression adjusted for demographic and clinical covariates. NSFR: Niacin skin flushing response; PND: Perinatal depression; DD: Depressive disorder.

Analysis of clinical symptoms, NSFR, and related factors within the PND group

Within the PND group, NSFR intensity a marked reduction was observed in the moderate depression subgroup as opposed to the severe depression subgroup (P = 0.024, Figure 2A). Conversely, the severe depression subgroup demonstrated a significantly elevated NSFR vs the combined mild/moderate subgroups (P = 0.035, Figure 2B). NSFR did not differ significantly across perinatal stages (first, second, or third trimester, and postpartum), with no statistically significant variation observed between periods. Postpartum depression severity showed a negative correlation with NSFR intensity (r = -0.539, P = 0.014, Figure 2C). In the PND group, anxiety symptom severity increased with maternal age (r = 0.357, P = 0.006, Figure 2D), a correlation not seen within the DD group. Both depressive (r = 0.342, P = 0.033, Figure 2E) and anxiety symptom severity (r = 0.313, P = 0.041, Figure 2F) displayed a direct proportional relationship with advancing gestational age. Furthermore, depressive (r = 0.704, P = 0.004, Figure 2G) and anxiety symptoms (r = 0.684, P = 0.004, Figure 2H) were positively associated with the time elapsed since delivery at postpartum follow-up visits. Finally, depressive and anxiety symptom scores displayed a statistically significant rise in the postpartum period than during pregnancy (P = 0.005 and P = 0.022, respectively; Figure 2I and J).

Figure 2 Analysis of clinical symptoms, niacin skin flushing response, and related factors in the perinatal depression group. A-J: Between-group comparisons were conducted using the Mann-Whitney U test, and correlations were evaluated using Spearman rank correlation analysis. PND: Perinatal depression; NSFR: Niacin skin flushing response; GAD-7: Generalized Anxiety Disorder 7-Item Scale; PHQ-9: Patient Health Questionnaire-9; HAMA: Hamilton Anxiety Rating Scale; HAMD-24: 24-Item Hamilton Depression Rating Scale.

Correlation analysis of diet and symptoms within the PND group

In comparison with the DD group, PND patients reported higher intakes of whole grains, freshwater fish, seafood, dark-colored vegetables, nuts, fruits, and liquid dairy products, and lower intakes of sugary foods and beer (Table 2). No significant differences were found for other dietary categories. Within the PND group, dark-colored vegetable intake correlated inversely with postpartum depression severity (r = -0.479, P = 0.044, Figure 3A), as did light-colored vegetable intake (r = -0.553, P = 0.017, Figure 3B) and egg intake (r = -0.852, P < 0.001, Figure 3C). Poultry intake correlated positively with depression severity (r = 0.358, P = 0.008, Figure 3D) and anxiety severity (r = 0.341, P = 0.011, Figure 3E).

Figure 3 Correlation analysis between dietary intake and symptom severity within the perinatal depression group. A-E: Spearman’s rank correlation test was used for correlation analysis. EPDS: Edinburgh Postnatal Depression Scale; HAMA: Hamilton Anxiety Rating Scale; HAMD-24: 24-Item Hamilton Depression Rating Scale.

Table 2 Comparison of dietary intake between the perinatal depression group and the depressive disorder group, mean ± SD.

	PND (g/day)	DD (g/day)	P value
Whole grains	43.2 ± 42.7	23.9 ± 34.9	< 0.001a
Freshwater aquatic products	40.6 ± 39	21.7 ± 22	0.005a
Marine aquatic products	31.1 ± 34.3	19.4 ± 32.2	0.003a
Dark-colored vegetables	163.8 ± 113.3	115.5 ± 107	0.006a
Nuts	27.2 ± 39.8	13.1 ± 20.4	0.010a
Fruits	198.3 ± 129.8	106.5 ± 89.1	< 0.001a
Liquid dairy products	172.8 ± 157	106.6 ± 97.1	0.020a
Sugary foods	52.5 ± 62.5	107.7 ± 129.4	0.101
Beer	22.5 ± 47.6	46.6 ± 80	0.002a

^aP < 0.05.

PND: Perinatal depression; DD: Depressive disorder.

DISCUSSION

The essential observations of the present work demonstrated that the NSFR was markable higher in patients with PND than in those with non-PND. This conclusion differs from most previous studies, which have generally reported a blunted NSFR in patients with major DD[11,12]. The exaggerated NSFR observed in patients with PND suggests a distinct underlying pathological mechanism closely linked to the unique neuroendocrine and immune disturbances during the perinatal period. Fluctuations in estrogen and progesterone levels may upregulate the expression of GPR109A in cutaneous microvascular endothelial cells or enhance the signaling sensitivity of this receptor, thereby amplifying the prostaglandin-mediated vasodilatory effect induced by niacin[10]. Meanwhile, existing studies have confirmed that the prostaglandin signaling pathway downstream of GPR109A can be regulated by endogenous hormonal status and inflammatory milieu, which further supports the mechanistic inference of the present study[10].

Notably, NSFR confer multiple clinical application values as an “objective auxiliary biomarker”. First, among women with first-onset depression during pregnancy or the postpartum period, NSFR helps distinguish PND from non-PND, especially in clinical scenarios where symptomatic overlap renders differential diagnosis based solely on clinical manifestations difficult. Unlike self-rating scales such as the EPDS and PHQ-9, NSFR is a physiological measurement index that is free from recall bias and social desirability effects. Second, NSFR presents obvious phase-specific variation characteristics: The response is elevated in patients with severe antenatal depression but relatively attenuated in those with postpartum depression. This feature can identify the transition from physiological compensation to decompensation, and is expected to provide a reference for determining the optimal timing of clinical intervention. Finally, given that the activity of the prostaglandin pathway changes alongside therapeutic responses, NSFR possesses potential research value as a dynamic indicator for monitoring disease progression. Future studies are warranted to establish the normal reference range of NSFR, evaluate its diagnostic efficiency by comparison with conventional screening scales, and further validate its application value in long-term follow-up monitoring.

Remarkably, within the PND group, patients with severe depression exhibited significantly higher NSFR than those with mild and moderate depression. However, this pattern was completely reversed in the postpartum subgroup: The more severe the depressive symptoms, the lower the NSFR level. In addition, no significant differences in NSFR levels were observed across different perinatal stages, including the first, second, and third trimesters as well as the postpartum period. These findings suggest that the association between NSFR and clinical symptoms of PND is complex and presents marked “stage dependence”. The enhanced NSFR in patients with severe PND may reflect a state of acute inflammatory activation or stress response, which further amplifies the skin flush effect mediated by the GPR109A signaling pathway[10]. This represents a unique biological characteristic under the specific neuroendocrine and immune physiological milieu of the perinatal period[13]. In contrast, the postpartum period is characterized by a sharp decline in hormone levels and immune system remodeling, accompanied by drastic fluctuations in reproductive hormones and a persistent high-grade proinflammatory state[13]. Such alterations in the postpartum inflammatory environment may exhaust the compensatory mechanisms that maintain high NSFR reactivity during pregnancy, ultimately leading to the downregulation of the function of skin flush-related signaling pathways.

This pattern of changes conforms to the framework of the “allostatic load theory”[25], which explains the process by which physiological systems gradually shift from adaptive regulation to decompensation under sustained stress. In PND, the initial elevation of the NSFR can be regarded as a “compensatory response” of the body to perinatal physiological stress. With the progression or aggravation of the condition, especially after entering the postpartum period, the physiological system transitions into a decompensated state, accompanied by a subsequent decline in NSFR levels. At this stage, classical pathogenetic mechanisms of depression, such as neurotransmitter depletion and chronic low-grade inflammation, become predominant. As depressive symptoms worsen, the NSFR decreases progressively. This also accounts for the characteristic pattern in postpartum depression whereby “more severe symptoms are accompanied by a lower NSFR”, a phenomenon consistent with findings in other psychiatric disorders. In summary, the present findings indicate that dynamic alterations in NSFR are jointly modulated by depression severity and perinatal stage, among which drastic postpartum hormonal fluctuations exert the most pronounced influence. PND occurs within a specific physiological time window, and its underlying pathogenesis is fundamentally distinct from that of non-PND.

In addition to differences in physiological indicators, the present study also identified characteristic dietary patterns in patients with PND. The overall dietary structure of the perinatal group was healthier, which is consistent with the increased awareness of nutritional health among women during pregnancy and the puerperium. Within the PND group, the intake of vegetables and eggs was inversely related to the severity of depressive symptoms, whereas higher intake of poultry meat was associated with more severe depression. Such correlations were not observed in the non-PND group. Constrained by the study design, it remains unclear whether the perinatal physiological state amplifies individual sensitivity to the above dietary factors; nevertheless, the relevant mechanistic inferences are biologically plausible. Vegetables are rich in folate, dietary fiber, vitamin C, flavonoids, and other antioxidants, which can alleviate emotional disturbances by regulating neurotransmitter synthesis and neuroinflammation[14], while also benefiting fetal neurodevelopment and maintaining maternal cholinergic homeostasis. Eggs serve as major dietary sources of choline and tryptophan. As a precursor of serotonin synthesis, tryptophan, together with choline, is closely involved in emotional regulation[26]. The positive correlation between poultry consumption and depressive severity may be attributed to its high content of branched-chain amino acids. Branched-chain amino acids vie with tryptophan for carriermediated transport across the blood-brain barrier, thereby reducing central serotonin synthesis[27]. Perinatal populations may be more susceptible to the metabolic interference of this pathway. All the above mechanistic speculations are based on existing literature. Further validation studies are warranted by detecting related nutrients, metabolites, and inflammatory biomarkers in this population.

NSFR was positively correlated with the intake of dark-colored vegetables, fruits, and dairy products, all of which are rich in polyunsaturated fatty acids (PUFAs). Serving as fundamental building blocks for cell membrane phospholipids, PUFAs can regulate the basal level of the systemic NSFR[28]. Therefore, NSFR intensity can not only reflect the activity of the niacin pathway but also mirror the overall nutritional status and immune-inflammatory profile during the perinatal period[28-30]. The dietary associations observed in patients with PND in this study may be partially mediated by the regulation of NSFR-related physiological pathways, which in turn interact with the disease progression of PND.

The mean age of participants in the PND group exceeded that of that in the non-PND group. Within the PND group, the severity of anxiety increased with maternal age, which is consistent with the findings of most previous studies. Advanced maternal age (≥ 35 years) has been well recognized as a risk factor for PND[31,32]. Women of advanced maternal age are more vulnerable to obstetric complications and bear heavier psychosocial burdens, including concerns about childbirth and childcare, workplace pressure, and worries associated with delayed childbearing[33]. The present study also found that increased gestational age, the postpartum period relative to pregnancy, and delayed medical consultation after delivery were all correlated with aggravated depressive symptoms, suggesting a “cumulative risk effect” of these factors. Postpartum hormonal decline, transformation of social roles, and sleep disturbance may collectively induce abnormalities in emotional regulation during this period[8].

Our study has several inherent limitations. As a cross-sectional investigation, it cannot establish causal relationships among variables. The associations of NSFR and dietary patterns with depressive severity may be bidirectional and subject to unmeasured confounding factors. Although participants with dermatological or immune disorders were excluded, baseline cutaneous vascular reactivity was not uniformly evaluated, which might introduce bias into NSFR quantification. A self-administered food frequency questionnaire was used to capture dietary information, inevitably leading to recall bias and social desirability bias, with perinatal women tending to overreport healthy food intake. Parallel measurements of reproductive hormones, serum fatty acids, inflammatory cytokines and other biochemical markers were not performed, making it difficult to elucidate the underlying mechanistic links among NSFR, nutritional status and perinatal physiological alterations. Furthermore, this investigation was limited to a single clinical site restricted to urban residents in Shanghai, limiting the generalizability of the findings to rural populations, other ethnic groups and diverse healthcare settings. In addition, the non-perinatal DD group had a relatively longer disease duration; despite all participants being first-episode and treatment-naive, chronic disease course might still interfere with NSFR outcomes. The sample sizes of antenatal and postpartum subgroups within the PND group were unbalanced, prohibiting robust stratified comparisons across gestational trimesters. Moreover, NSFR was tested only once for each participant, leaving test-retest reliability and day-to-day variability unassessed.

To address these shortcomings, future multicenter prospective cohort studies are warranted. Such studies should incorporate multidimensional objective biomarkers, including reproductive hormones, plasma tryptophan and branched-chain amino acid levels, PUFA profiles, as well as inflammatory markers such as C-reactive protein, interleukin-6 and tumor necrosis factor alpha, while enlarging the sample size and enrolling participants with more diverse demographic characteristics. By optimizing study design and laboratory assessment protocols, further research is expected to validate the mechanistic cascade of “diet-biochemical profiles-NSFR-clinical symptoms”. This will overcome the inherent limitation of cross-sectional studies confined to correlation analysis, and clarify the causal relationships and regulatory mechanisms among relevant influencing factors.

CONCLUSION

This investigation is the earliest to systematically characterize the phenotype of the NSFR in patients with PND and its association with nutritional factors. Two core findings were identified: Patients with PND exhibited a characteristic exaggerated NSFR response distinct from that of non-PND; furthermore, the association between NSFR and the severity of depressive symptoms presents perinatal stage specificity and is concurrently modulated by dietary factors. The research findings confirm that PND possesses unique perinatal-specific pathophysiological mechanisms, indicating the necessity of classifying it as an independent clinical disease entity. The dietary-related findings of this study suggest that increased intake of vegetables and eggs as well as reduced consumption of poultry may be associated with decreased symptom severity of PND, while the causal nature of these associations remains to be further verified. NSFR holds promise as an objective biomarker to differentiate PND from other depressive subtypes, and can also reflect systemic nutritional and metabolic status. Future prospective studies incorporating the detection of biochemical indicators are still needed to validate its clinical application value. Risk factors identified for PND onset and worsening include advanced maternal age, later gestational stage, the postpartum period, and delayed care-seeking. These findings support routine mental health screening in older pregnant women, with particular attention during the third trimester and postpartum. Early identification and intervention, especially postpartum, may reduce the risk of symptom chronicity.