Prevalence and associated factors of psychotic symptoms in young first-episode, drug-naïve major depressive disorder patients with abnormal lipid metabolism Authors: Boxuan Li, Minghui Li, Mengqian Li, Xing Wang, Mingxiu Kang, Guoshuang Zhang, Guangming Xu, Xiangyang Zhang Published: 2025 ABSTRACT AbstractBackgroundComorbidity of major depressive disorder (MDD) and abnormal lipid metabolism (ALM) is common. Psychotic major depression (PMD) is a severe subtype of MDD. However, the occurrence of psychotic symptoms in MDD patients with ALM has not been thoroughly investigated. This study aimed to explore the risk factors for psychotic symptoms in MDD patients with ALM.MethodsA total of 1289 first-episode, drug-naïve (FEDN) MDD patients aged 18–45 years were recruited. Sociodemographic data and various clinical parameters, including lipid profiles, thyroid function tests, and thyroid antibodies, were measured. Patients were evaluated using the Hamilton Rating Scale for Depression (HAMD), Hamilton Anxiety Rating Scale (HAMA), and the positive subscale of the Positive and Negative Syndrome Scale (PANSS).ResultsIn the FEDN MDD sample, the prevalence of ALM was 81.12% (1047/1289). Among MDD patients with ALM, levels of total cholesterol (TC), HAMA, HAMD, suicide attempts, TSH, TgAb, TPOAb, fasting glucose, and blood pressure were significantly higher in the PMD group than in the NPMD group. Binary logistic regression indicated that HAMD scores (OR = 1.187 [1.045, 1.349], P = 0.0085), HAMA scores (OR = 1.699 [1.510, 1.912], P < 0.0001), and TSH levels (OR = 1.116 [1.001, 1.234], P = 0.0472) were independent risk factors for psychotic symptoms in MDD patients with ALM.ConclusionOur study identified a high prevalence of psychotic symptoms in young MDD patients with ALM. Key predictors, including HAMD, HAMA, and TSH, demonstrated strong predictive accuracy. These findings highlight the potential of these factors for early identification of psychotic symptoms in this patient population. Future research is needed to validate these results and explore the underlying mechanisms.Keywords: Abnormal lipid metabolism, Psychotic major depression, Major depressive disorder, Cross-sectional study, Clinical correlates Introduction Major depressive disorder (MDD) is a debilitating psychiatric condition characterized by core symptoms including persistent low mood, anhedonia, and cognitive dysfunction [1, 2]. According to the World Health Organization (WHO), the 12-month prevalence of major depressive disorder in the general population was approximately 4.4% [3], and the lifetime prevalence was 10–15% globally [4]. The prevalence of MDD is more prevalent among young adults than olds [5–7]. The disorder imposes substantial societal burden, accounting for 37.2 million years lived with disability (YLDs) in 2019—representing 29.7% of total mental disorder-related YLDs [8]. Notably, over 60% of MDD patients experience recurrent episodes, with treatment resistance occurring in 30–50% of cases, leading to functional impairment comparable to chronic somatic diseases [5]. The relationship between dyslipidemia and major depressive disorder (MDD) has emerged as a critical focus in psychosomatic research. Epidemiological evidence reveals a pronounced comorbidity pattern: 52.3% of MDD patients exhibit hypertriglyceridemia, while 44.7% present with hypercholesterolemia [9]. Notably, comparative analyses demonstrate significantly altered lipid profiles in MDD populations relative to healthy controls. Specifically, serum high-density lipoprotein cholesterol (HDL-C) levels show consistent reductions in depressed individuals, with parallel observations of elevated low-density lipoprotein cholesterol (LDL-C) concentrations and an increased LDL/HDL ratio [10, 11]. Beyond simple correlation, specific lipid components are linked to the clinical dynamics of depression. For instance, higher LDL-C levels have been found to correlate not only with the presence of depression but also with its severity and prospective short-term course, suggesting a role for lipid metabolism in the pathophysiology of depressive states [12]. This metabolic dysregulation pattern suggests potential bidirectional pathways linking lipid homeostasis and neuropsychiatric pathophysiology. Furthermore, recent evidence indicates that this link extends beyond mood to include cognitive functions. Studies have shown that the protective effect of HDL on cognition may be diminished or absent in individuals with depressive symptoms, highlighting a complex interplay where depression mediates the relationship between lipid metabolism and cognitive health [13]. Further metabolic investigations identify a characteristic triad in chronically depressed patients undergoing antidepressant therapy: depressed HDL-C levels, hypertriglyceridemia, and hyperglycemia [14]. This metabolic dysregulation pattern suggests potential bidirectional pathways linking lipid homeostasis and neuropsychiatric pathophysiology. Particularly relevant are intervention studies examining modifiable risk factors. A randomized trial demonstrated that 12-week aerobic exercise interventions (walking/bicycling) produced simultaneous improvements in both lipid profiles and depressive symptomatology, achieving 15–22% reductions in LDL-C and triglycerides, alongside 18% HDL-C elevation and significant Beck Depression Inventory score decreases [15]. Psychotic major depression (PMD) represents a severe subtype of major depressive disorder (MDD) characterized by psychotic features such as delusions and hallucinations, distinguishing it from nonpsychotic major depression (NPMD) in terms of diagnostic criteria, treatment approaches, and prognosis [16]. Studies indicate that psychotic symptoms are present in a significant proportion of MDD patients, with prevalence rates of 10.95% in middle-aged adults [17] and 25.3% across a broader MDD population [18]. Compared with NPMD, PMD is associated with more severe symptoms [19, 20], poorer prognosis [21, 22], increased recurrence risk [23, 24], higher mortality rates [25], and elevated levels of anxiety and depression [26]. The onset of psychotic symptoms in MDD is driven by a complex interplay of neurobiological, clinical, and biochemical factors. Neurobiologically, neurotransmitter dysregulation (involving dopamine, serotonin, glutamate, and GABA), neuroinflammation, oxidative stress, and mitochondrial dysfunction play crucial roles in the pathophysiology of psychosis [27]. Clinically, the severity of anxiety and depression, as assessed by tools like the Hamilton Anxiety Rating Scale (HAMA) and the Hamilton Depression Rating Scale (HAMD), is a strong predictor of psychotic symptomatology. Additionally, biochemical factors such as thyroid-stimulating hormone (TSH) levels and metabolic markers like cholesterol and triglycerides have been linked to the development of psychosis [17, 28]. Patients with MDD exhibiting psychotic features are at a higher risk for severe anxiety and suicidal tendencies, emphasizing the need for vigilant monitoring and management of these patients [25]. Therefore, a more in-depth understanding of the risk factors and clinical characteristics of PMD is essential to develop improved treatment strategies. Accumulating evidence highlights a clinically significant comorbidity between major depressive disorder (MDD) and dyslipidemia, particularly in patients with persistent metabolic abnormalities. Epidemiological studies reveal that over half of MDD patients (52.3%) present with hypertriglyceridemia, while 44.7% concurrently exhibit hypercholesterolemia [9], suggesting lipid dysregulation may constitute a distinct pathophysiological subtype. Cross-sectional comparisons further demonstrate characteristic lipid profile deviations: MDD patients show 23% lower serum HDL-C levels (P < 0.01) versus healthy controls, accompanied by elevated LDL-C concentrations and a pro-atherogenic LDL/HDL ratio exceeding 3.5 [11]. Notably, chronic MDD populations under antidepressant treatment exhibit exacerbated metabolic disturbances, with 68% displaying a concurrent triad of hypohdlipoproteinemia (HDL-C < 40 mg/dL), hypertriglyceridemia (>150 mg/dL), and insulin resistance (HOMA-IR > 2.6) [10]. This persistent dysmetabolic phenotype not only correlates with 2.1-fold higher risks of treatment resistance (95% CI 1.4–3.2) [14], but also implies shared biological mechanisms between lipid homeostasis and mood regulation. Young adults are highly susceptible to mental disorders that diminish quality of life and are a leading global cause of disability. Their health and well-being are significantly compromised by mental disorders [29]. Young adults tend to exhibit more prominent psychotic symptoms than older adults [30]. A well-established inverse relationship exists between positive psychotic symptoms and age [31]. In the general population, psychotic experiences are most prevalent in childhood and decline steadily from adolescence through adulthood and old age [32]. However, to the best of our knowledge, no studies have investigated the prevalence and risk factors for psychotic symptoms in a large group of young MDD patients with abnormal lipid metabolism. The current study aimed to (1) determine the prevalence of psychotic symptoms in young First-Episode and Drug-Naïve (FEDN) major depressive disorder (MDD) patients, and (2) explore the risk factors for the development of psychotic symptoms in this population. Materials and methods Sample Between 2015 and 2017, the current study was conducted at a psychiatric outpatient department of a general hospital in Taiyuan, Shanxi Province. This study enrolled 1047 participants who met a series of predefined inclusion and exclusion criteria. To be included, individuals were required to: (1) be of Han Chinese descent; (2) be between 18 and 45 years of age; (3) have a primary diagnosis of major depressive disorder (MDD) confirmed according to the Diagnostic and Statistical Manual of Mental Disorders, 4th Edition, Text Revision (DSM-IV-TR); and (4) exhibit a baseline severity score of ≥24 on the 17-item Hamilton Depression Rating Scale (HAMD-17). A critical criterion for inclusion was that all participants were experiencing their first episode of MDD and were treatment-naïve, defined as having no prior history of systematic treatment with antidepressants, antipsychotics, mood stabilizers, or anxiolytics. Furthermore, participants were required to be present with dyslipidemia, characterized by at least one of the following profiles: total cholesterol (TC) > 5.17 mmol/L, triglycerides (TG) > 1.7 mmol/L, high-density lipoprotein cholesterol (HDL-c) < 1.04 mmol/L, or low-density lipoprotein cholesterol (LDL-c) > 3.12 mmol/L. Exclusion criteria included a current or past diagnosis of alcohol or substance use disorder (nicotine use was permitted), pregnancy or lactation, and the presence of any severe, unstable, or life-threatening medical conditions, such as central nervous system disorders or cancer. The Institutional Review Board of the First Clinical Medical College of Shanxi Medical University approved this research. Participants in this study provided informed consent by signing a consent form. Socioeconomic factors and clinical measurements Well-trained researchers collected demographic and clinical information for the study. Each patient completed a detailed questionnaire that included general information as well as sociodemographic characteristics. In addition, we gathered information from available medical records. The 17-item Hamilton Assessment of Depression Scale (HAMD) was used to comprehensively evaluate the severity of depressive symptoms [33]. Eight of the 17 scale items use a 5-point scale ranging from 0 to 4 (absent to severe), while the remaining nine use a 3-point scale ranging from 0 to 2 (missing to specify the severity of symptoms). The 14-item Hamilton Anxiety Scale (HAMA) was used to comprehensively evaluate the severity of anxiety symptoms [34]. All of the 14 scale items use a 5-point scale ranging from 0 to 4, where 0 represents “absent” and 4 represents “extremely severe”. The Positive and Negative Syndrome Scale (PANSS) positive symptom subscale was used to assess the severity of psychotic symptoms [35]. Each of the seven subscale items on the PANSS uses a seven-point Likert scale. The group was divided based on a threshold score of 15 to determine whether they were psychotic. Young patients with psychotic symptoms and depression should have a PANSS score of 15 or higher [25]. Blood samples All participants provided serum samples between the hours of 6:00 a.m. and 8:00 a.m., following an overnight fast. By 11 a.m. on the same day, all blood samples were sent had been shipped to the hospital laboratory for analysis. During the fasting period, the following biochemicals were measured: blood glucose, TC, triglycerides (TG), HDL, and LDL, as well as thyroid stimulating hormone (TSH), free triiodothyronine 3 (FT3), free thyroxine 4 (FT4), thyroid peroxidase antibody (TPOAb), and anti-thyroglobulin (TgAb). Statistical analysis All data were anonymized and processed in compliance with the Declaration of Helsinki. Continuous variables were first assessed for normality using the Kolmogorov–Smirnov test. Between-group comparisons were performed as follows: normally distributed variables were analyzed using ANOVA, non-normally distributed variables with the Mann–Whitney U test, and categorical variables with Pearson’s chi-square test. To address multicollinearity and optimize variable selection, least absolute shrinkage and selection operator (LASSO) regression with tenfold cross-validation was first applied to identify key predictors, followed by multivariable logistic regression analysis on selected variables. Odds ratios (ORs) with 95% confidence intervals (CIs) were calculated, and model performance was evaluated via receiver operating characteristic (ROC) curve analysis. We applied Bonferroni correction to all univariate analyses, adopting a corrected significance threshold of P < 0.05. All analyses were performed using SAS 9.4, with statistical significance defined as P < 0.05 (two-tailed). Results Prevalence of psychotic symptoms of young MDD patients with and without abnormal lipid metabolism The proportion of abnormal lipid metabolism in young patients was 81.12% (1047/1289). As presented in Table 1, the rate of psychotic symptoms in the ALM group was nearly four times that of the non-ALM group (9.65% vs. 2.48%, P = 0.0069).The proportion of suicidal attempts was higher in the ALM group (20.53% vs. 11.98%, P = 0.0506). The patients with ALM also had higher median HAMD (31 [29, 32] vs. 28 [26, 30], P < 0.0001) and HAMA (21 [18, 23] vs. 19 [17, 22], P < 0.0001) scores. In addition, patients with ALM had higher median TSH (5.24 [3.32, 6.84] μIU/L vs. 3.62 [2.14, 4.47] μIU/L, P < 0.0001), TGAb (21.57 [14.43, 43.50] IU/L vs. 19.27 [13.50, 29.79] IU/L, P = 0.8096), while there was no significant difference in TPOAb (17.13 [12.30, 33.65] IU/L vs. 17.24 [12.53, 28.16] IU/L, P > 1). For thyroid-related hormone FT3 and FT4, there were no significant differences (FT3: 4.97 [4.39, 5.46] pmol/L vs. 4.77 [4.32, 5.34] pmol/L, P > 1; FT4: 16.58 [14.58, 18.80] pmol/L vs. 16.61 [14.30, 18.70] pmol/L, P > 1). Table 1Sample characteristics of young patients with and without abnormal lipid metabolismVariableOverallN = 1289With ALMN = 1047Without ALMN = 242Chi sq/ZBonferroni-corrected P valueaAge, median (IQR)28 (21, 37)28 (21, 38)27 (21, 35)−1.40091.0000Gender1.00381.0000 Male, N (%)470 (36.46)375 (35.82)95 (39.26) Female, N (%)819 (63.54)672 (64.18)147 (60.74)Education level0.00201.0000 Middle school, N (%)792 (61.44)643 (61.41)149 (61.57) College degree or above, N (%)497 (38.56)404 (38.59)93 (38.43)Marital status2.41771.0000 Unmarried, N (%)487 (37.78)385 (36.77)102 (42.15) Married, N (%)802 (62.22)662 (63.23)140 (57.85)BMI, median (IQR) (kg/m2)24.22 (23.20, 25.60)24.25 (23.21, 25.80)24.12 (23.12, 25.21)−2.68080.1679Illness duration, median (IQR) (m)4.5 (3.0, 7.0)5 (3, 7)4 (2, 6)−2.29120.5060Age at onset, median (IQR) (years)28 (21, 37)28 (21, 37)27 (20, 35)−1.43421.0000Psychotic symptoms13.26540.0069 No, N (%)1182 (91.70)236 (97.52)946 (90.35) Yes, N (%)107 (8.30)6 (2.48)101 (9.65)TC, mean ± SD (mmol/L)5.16 (4.43, 5.92)5.38 (4.62, 6.16)4.38 (3.86, 4.78)−15.6615<0.0001TG, median (IQR) (mmol/L)1.97 (1.38, 2.78)2.26 (1.70, 2.88)1.26 (1.07, 1.47)−19.1425<0.0001HDL-c, median (IQR) (mmol/L)1.24 (1.03, 1.42)1.23 (0.97, 1.41)1.28 (1.20, 1.52)6.6544<0.0001LDL-c, median (IQR) (mmol/L)2.90 (2.30, 3.50)3.10 (2.44, 3.62)2.50 (2.12, 2.84)−10.5<0.0001Suicidal attempted9.36660.0506 No, N (%)1045 (81.07)832 (79.47)213 (88.02) Yes, N (%)244 (18.93)215 (20.53)29 (11.98)HAMD30 (28, 32)31 (29, 32)28 (26, 30)−11.6189<0.0001HAMA21 (18, 23)21 (18, 23)19 (17, 22)−6.0866<0.0001TSH, median (IQR) (μIU/L)4.73 (3.04, 6.52)5.24 (3.32, 6.84)3.62 (2.14, 4.47)−10.4123<0.0001FT3, median (IQR) (pmol/L)4.94 (4.38, 5.44)4.97 (4.39, 5.46)4.77 (4.32, 5.34)−1.86121.0000FT4, median (IQR) (pmol/L)16.58 (14.56, 18.77)16.58 (14.58, 18.80)16.61 (14.30, 18.70)−0.71761.0000TgAb, median (IQR) (IU/L)21.24 (14.25, 37.76)21.57 (14.43, 43.50)19.27 (13.50, 29.79)−2.10640.8096TPOAb, median (IQR) (IU/L)17.22 (12.32, 32.72)17.13 (12.30, 33.65)17.24 (12.53, 28.16)−0.98821.0000Fasting glucose, median (IQR) (mmol/L)5.32 (4.92, 5.77)5.38 (4.95, 5.82)5.12 (4.85, 5.51)−5.0188<0.0001SBP, median (IQR) (mmHg)118 (110, 124)118 (110, 124)112 (108, 120)−6.6184<0.0001DBP, median (IQR) (mmHg)75 (70, 80)76 (71, 80)74 (70, 78)−5.6853<0.0001ALM abnormal lipid metabolism, BMI body mass index, TC total cholesterol, TG triglyceride, HDL-c high-density lipoprotein cholesterol, LDL-c low-density lipoprotein cholesterol, HAMD Hamilton Assessment of Depression Scale, HAMA Hamilton Assessment of Anxiety Scale, TSH thyroid stimulating hormone, FT3 free triiodothyronine 3, FT4 free thyroxine 4, TgAb anti-thyroglobulin, TPOAb thyroid peroxidase antibody, SBP systolic blood pressure, DBP diastolic blood pressureaBonferroni-corrected P > 1 is shown as P = 1 Sample characteristics of young patients with and without abnormal lipid metabolism ALM abnormal lipid metabolism, BMI body mass index, TC total cholesterol, TG triglyceride, HDL-c high-density lipoprotein cholesterol, LDL-c low-density lipoprotein cholesterol, HAMD Hamilton Assessment of Depression Scale, HAMA Hamilton Assessment of Anxiety Scale, TSH thyroid stimulating hormone, FT3 free triiodothyronine 3, FT4 free thyroxine 4, TgAb anti-thyroglobulin, TPOAb thyroid peroxidase antibody, SBP systolic blood pressure, DBP diastolic blood pressure aBonferroni-corrected P > 1 is shown as P = 1 Differences of psychotic symptoms vs. nonpsychotic symptoms in young MDD patients with abnormal lipid metabolism As shown in Table 2, among young patients with abnormal lipid metabolism, the proportion of those with psychotic symptoms (PMD) was 101/1047 (9.65%). There was a significant difference in age between the two groups, with the median age in the PMD group being 26 (21, 33) years and 29 (21, 38) years in the NPMD group (P > 1). Regarding lipid-related indicators, the PMD group had a higher median TC (5.73 [5.02, 6.34] mmol/L vs. 5.35 [4.60, 6.14] mmol/L, P = 0.0374), and a similar TG (2.50 [1.86, 3.10] mmol/L vs. 2.25 [1.68, 2.87] mmol/L, P > 1), HDL-c (1.19 [0.95, 1.35] mmol/L vs. 1.23 [0.97, 1.41] mmol/L, P > 1), LDL-c (3.21 [2.47, 3.70] mmol/L vs. 3.10 [2.82, 3.60] mmol/L, P > 1). Table 2Sample characteristics of young patients with abnormal lipid metabolism with and without psychotic symptomsVariableOverallN = 1047PMDN = 101NPMDN = 946Chi sq/ZBonferroni-corrected P valueaAge, median (IQR)28 (21, 38)26 (21, 33)29 (21, 38)−1.9711.0000Gender1.81751.0000 Male, N (%)375 (35.82)30 (29.70)345 (36.47) Female, N (%)672 (64.18)71 (70.30)601 (63.53)Education level0.04881.0000 Middle school, N (%)643 (61.41)61 (60.40)582 (61.52) College degree or above, N (%)404 (38.59)40 (39.60)364 (38.48)Marital status2.21831.0000 Unmarried, N (%)385 (36.77)44 (43.56)341 (36.05) Married, N (%)662 (63.23)57 (56.44)605 (63.95)BMI, median (IQR) (kg/m2)24.25 (23.21, 25.80)24.56 (23.24, 26.12)24.23 (23.21, 25.63)0.86441.0000Illness duration, median (IQR) (m)5.00 (3.00, 7.00)5.00 (2.50, 7.00)4.75 (3.00, 8.00)−0.78651.0000Age at onset, median (IQR) (years)28 (21, 37)26 (21, 33)28 (21, 37)−1.92721.0000TC, mean ± SD (mmol/L)5.38 (4.62, 6.16)5.73 (5.02, 6.34)5.35 (4.60, 6.14)3.13070.0374TG, median (IQR) (mmol/L)2.26 (1.70, 2.88)2.50 (1.86, 3.10)2.25 (1.68, 2.87)1.74131.0000HDL-c, median (IQR) (mmol/L)1.23 (0.97, 1.41)1.19 (0.95, 1.35)1.23 (0.97, 1.41)−0.95731.0000LDL-c, median (IQR) (mmol/L)3.10 (2.44, 3.62)3.21 (2.70, 3.70)3.10 (2.40, 3.60)1.81821.0000Suicidal attempted46.3072<0.0001 No, N (%)832 (79.47)54 (53.47)778 (82.24) Yes, N (%)215 (20.53)47 (46.53)168 (17.76)HAMD31 (29, 32)34 (32, 36)30 (28, 32)10.9524<0.0001HAMA21 (18, 23)27 (24, 28)20 (18, 22)13.6497<0.0001TSH, median (IQR) (μIU/L)5.24 (3.32, 6.84)7.01 (5.58, 9.48)5.05 (3.19, 6.62)7.5264<0.0001FT3, median (IQR) (pmol/L)4.97 (4.39, 5.46)4.86 (4.33, 5.41)4.98 (4.41, 5.47)−0.8171.0000FT4, median (IQR) (pmol/L)16.58 (14.58, 18.80)16.63 (15.16, 18.70)16.54 (14.58, 18.80)0.51491.0000TgAb, median (IQR) (IU/L)21.57 (14.43, 43.50)23.21 (15.60, 118.23)21.27 (14.33, 39.83)2.07420.8382TPOAb, median (IQR) (IU/L)17.13 (12.30, 33.65)21.19 (14.13, 54.25)16.78 (12.25, 32.86)2.80810.1100Fasting glucose, median (IQR) (mmol/L)5.38 (4.95, 5.82)5.54 (5.18, 6.07)5.36 (4.94, 5.79)3.28290.022SBP, median (IQR) (mmHg)118 (110, 124)123 (116, 128)118 (110, 124)3.60790.0066DBP, median (IQR) (mmHg)76 (71, 80)78 (74, 82)75 (70, 80)3.88220.0022PMD psychotic major depression, NPMD nonpsychotic depression, BMI body mass index, TC total cholesterol, TG triglyceride, HDL-c high-density lipoprotein cholesterol, LDL-c low-density lipoprotein cholesterol, HAMD Hamilton Assessment of Depression Scale, HAMA Hamilton Assessment of Anxiety Scale, TSH thyroid stimulating hormone, FT3 free triiodothyronine 3, FT4 free thyroxine 4, TgAb anti-thyroglobulin, TPOAb thyroid peroxidase antibody, SBP systolic blood pressure, DBP diastolic blood pressureaBonferroni-corrected P > 1 is shown as P = 1 Sample characteristics of young patients with abnormal lipid metabolism with and without psychotic symptoms PMD psychotic major depression, NPMD nonpsychotic depression, BMI body mass index, TC total cholesterol, TG triglyceride, HDL-c high-density lipoprotein cholesterol, LDL-c low-density lipoprotein cholesterol, HAMD Hamilton Assessment of Depression Scale, HAMA Hamilton Assessment of Anxiety Scale, TSH thyroid stimulating hormone, FT3 free triiodothyronine 3, FT4 free thyroxine 4, TgAb anti-thyroglobulin, TPOAb thyroid peroxidase antibody, SBP systolic blood pressure, DBP diastolic blood pressure aBonferroni-corrected P > 1 is shown as P = 1 The proportion of suicidal attempts was higher in the PMD group (46.53% vs. 17.76%, P < 0.0001). The PMD group also had higher median HAMD (34 [32, 36] vs. 30 [28, 32], P < 0.0001) and HAMA (27 [24, 28] vs. 20 [18, 22], P < 0.0001) scores. The PMD group had a higher median TSH (7.01 [5.58, 9.48] μIU/L vs. 5.05 [3.19, 6.62] μIU/L, P < 0.0001), TGAb (23.21 [15.60, 118.23] IU/L vs. 21.27 [14.33, 39.83] IU/L, P = 0.8382), and TPOAb (21.19 [14.13, 54.25] IU/L vs. 16.78 [12.25, 32.86] IU/L, P = 0.1100). Differences of psychotic symptoms vs. nonpsychotic symptoms in young MDD patients without abnormal lipid metabolism Among the 242 young MDD patients without abnormal lipid metabolism, the prevalence of PMD was 2.48% (6/242). As detailed in Table 3, the proportion of suicidal attempts was significantly higher in the PMD group compared to the NPMD group (66.67% vs. 10.59%, P = 0.0462). Similarly, the PMD group exhibited significantly higher median HAMD scores (34 [32, 35] vs. 28 [26, 30], P < 0.0001) and HAMA scores (26 [25, 28] vs. 19 [17, 22], P < 0.0001). There were no statistically significant differences between the two groups in terms of demographic characteristics (age, gender, education level, marital status), BMI, or levels of TC, TG, HDL-c, LDL-c, and thyroid-related indicators. Table 3Sample characteristics of young patients without abnormal lipid metabolism with and without psychotic symptomsVariableOverallN = 242PMDN = 6NPMDN = 236Chi sq/ZBonferroni-corrected P valueaAge, median (IQR)28 (21, 35)30 (18, 39)28 (21, 35)0.08581.0000Gender1.0000 Male, N (%)95 (39.26)3 (50.00)92 (38.98) Female, N (%)147 (60.74)3 (50.00)144 (61.02)Education level–– Middle school, N (%)149 (61.57)6 (100.00)143 (60.59) College degree or above, N (%)93 (38.43)0 (0.00)93 (39.41)Marital status1.0000 Unmarried, N (%)102 (42.15)2 (33.33)100 (42.37) Married, N (%)140 (57.85)4 (66.67)136 (57.63)BMI, median (IQR) (kg/m2)24.12 (23.12, 25.21)23.34 (22.58, 23.98)24.13 (23.13, 25.22)−1.05711.0000Illness duration, median (IQR) (m)4.00 (2.50, 6.00)4.00 (3.50, 5.00)4.00 (2.50, 6.00)−0.31411.0000Age at onset, median (IQR) (years)28 (20, 35)30 (18, 39)28 (21, 35)0.12131.0000TC, mean ± SD (mmol/L)4.39 (3.86, 4.78)4.64 (4.60, 4.68)4.37 (3.86, 4.78)1.17231.0000TG, median (IQR) (mmol/L)1.26 (1.07, 1.47)1.44 (1.26, 1.50)1.26 (1.06, 1.47)1.04551.0000HDL-c, median (IQR) (mmol/L)1.29 (1.20, 1.52)1.22 (1.06, 1.26)1.29 (1.20, 1.52)−1.88741.0000LDL-c, median (IQR) (mmol/L)2.50 (2.12, 2.84)2.56 (2.30, 2.73)2.50 (2.12, 2.85)0.24221.0000Suicidal attempted0.0462 No, N (%)213 (88.02)2 (33.33)211 (89.41) Yes, N (%)29 (11.98)4 (66.67)25 (10.59)HAMD28 (26, 30)34 (32, 35)28 (26, 30)3.9000<0.0001HAMA19 (17, 22)26 (25, 28)19 (17, 22)3.9191<0.0001TSH, median (IQR) (μIU/L)3.63 (2.14, 4.47)5.11 (3.47, 9.40)3.63 (2.09, 4.56)1.76871.0000FT3, median (IQR) (pmol/L)4.77 (4.32, 5.34)4.64 (4.30, 5.63)4.77 (4.33, 5.34)−0.03841.0000FT4, median (IQR) (pmol/L)16.61 (14.30, 18.70)16.71 (15.67, 17.33)16.61 (14.26, 18.77)−0.09741.0000TgAb, median (IQR) (IU/L)19.27 (13.50, 29.79)19.53 (12.22, 25.44)19.27 (13.51, 29.86)−0.33071.0000TPOAb, median (IQR) (IU/L)17.24 (12.53, 28.16)12.75 (10.82, 15.73)17.37 (12.55, 28.22)−1.14271.0000Fasting glucose, median (IQR) (mmol/L)5.12 (4.85, 5.51)5.78 (5.29, 6.15)5.12 (4.85, 5.49)2.36830.3938SBP, median (IQR) (mmHg)112 (108, 120)118 (112, 120)112 (107, 120)1.24211.0000DBP, median (IQR) (mmHg)74 (70, 78)71 (68, 74)74 (70, 78)−0.79611.0000PMD psychotic major depression, NPMD nonpsychotic depression, BMI body mass index, TC total cholesterol, TG triglyceride, HDL-c high-density lipoprotein cholesterol, LDL-c low-density lipoprotein cholesterol, HAMD Hamilton Assessment of Depression Scale, HAMA Hamilton Assessment of Anxiety Scale, TSH thyroid stimulating hormone, FT3 free triiodothyronine 3, FT4 free thyroxine 4, TgAb anti-thyroglobulin, TPOAb thyroid peroxidase antibody, SBP systolic blood pressure, DBP diastolic blood pressureaBonferroni-corrected P > 1 is shown as P = 1 Sample characteristics of young patients without abnormal lipid metabolism with and without psychotic symptoms PMD psychotic major depression, NPMD nonpsychotic depression, BMI body mass index, TC total cholesterol, TG triglyceride, HDL-c high-density lipoprotein cholesterol, LDL-c low-density lipoprotein cholesterol, HAMD Hamilton Assessment of Depression Scale, HAMA Hamilton Assessment of Anxiety Scale, TSH thyroid stimulating hormone, FT3 free triiodothyronine 3, FT4 free thyroxine 4, TgAb anti-thyroglobulin, TPOAb thyroid peroxidase antibody, SBP systolic blood pressure, DBP diastolic blood pressure aBonferroni-corrected P > 1 is shown as P = 1 Factors associated with PMD in young FEDN MDD with abnormal lipid metabolism Lasso regression was initially employed to screen variables, resulting in the selection of HAMD, HAMA, and TSH as the final variables. Subsequently, four demographic information variables, namely gender, age, educational attainment, and marital status, were incorporated for variable-control purposes. The results of the binary logistic regression analysis showed that the risk factors for psychotic symptoms included HAMD (OR = 1.187, 95% CI = 1.045–1.349, P = 0.0085), HAMA (OR = 1.699, 95% CI = 1.510–1.912, P < 0.001) and TSH (OR = 1.116, 95% CI = 1.001–1.243, P = 0.0472) as shown in Table 4. Table 4Multivariate binary logistic regression for factors associated with psychotic symptoms in young FEDN MDD patients with lipid metabolism disordersVariableCoefficientsS.E.OR (95% CI)PHAMD0.17160.06521.187 (1.045, 1.349)0.0085HAMA0.53000.06031.699 (1.510, 1.912)<0.0001TSH (μIU/L)0.10950.05521.116 (1.001, 1.243)0.0472Adjusted for demographic information (age, gender, education, married status)HAMD Hamilton Assessment of Depression Scale, HAMA Hamilton Assessment of Anxiety Scale, TSH thyroid stimulating hormone Multivariate binary logistic regression for factors associated with psychotic symptoms in young FEDN MDD patients with lipid metabolism disorders Adjusted for demographic information (age, gender, education, married status) HAMD Hamilton Assessment of Depression Scale, HAMA Hamilton Assessment of Anxiety Scale, TSH thyroid stimulating hormone The areas under the ROC curve (AUC) for TSH, HAMD and HAMA were 0.7459, 0.0.8411, and 0.9199, respectively. In addition, a higher AUC of 0.9258 was achieved when the three risk factors were combined (Fig. 1). The ability of related factors to distinguish between PMD and NPMD in young MDD patients with abnormal lipid metabolism Discussion This study aims to investigate the prevalence of psychotic symptoms and their influencing factors among young MDD patients with abnormal lipid metabolism in China. Our major findings are as follows: First, the proportion of abnormal lipid metabolism in young patients reached 81.12% (1047/1289), and patients with abnormal lipid metabolism had significant differences in multiple lipid-related indicators compared with those with normal lipid metabolism. Second, among young MDD patients with abnormal lipid metabolism, factors associated with psychotic symptoms included HAMD, HAMA, and TSH, as determined by Lasso regression and binary logistic regression. Third, the areas under the ROC curve for TSH, HAMD, and HAMA in predicting psychotic symptoms were 0.7459, 0.8411, and 0.9199 respectively, and when these three factors were combined, a higher AUC of 0.9258 was achieved, indicating better prediction accuracy. Prevalence of abnormal lipid metabolism in young MDD patients In this study, the prevalence of abnormal lipid metabolism (ALM) among young MDD patients was notably high, reaching 81.12%. This finding is worthy of attention, as such a high prevalence can significantly exacerbate the risk of cardiovascular diseases and other metabolic complications [36]. This prevalence is consistent with findings from various studies that have explored the intersection of mental health disorders and metabolic dysfunctions. For instance, a study investigating the prevalence and factors associated with dyslipidemia in patients with first hospitalization for MDD found a high incidence of hyperlipidemia (HL) at 72.17% among these patients. This study highlighted that higher scores on the Hamilton Depression Scale (HAMD) and the Clinical Global Impression Scale-Severity of Illness (CGI-SI) were significant risk factors for co-morbid HL in MDD patients, suggesting a strong link between the severity of depression and lipid metabolism disorders [37]. Moreover, the research on the association between lipid levels and depression has shown that patients with major depression often exhibit adverse plasma lipid patterns, such as higher triglycerides (TG) and lower high-density lipoprotein (HDL) cholesterol levels compared to healthy controls. This dyslipidemic profile is associated with increased cardiovascular risk, underscoring the need for routine lipid measurements in depressive individuals to better manage their overall health and reduce potential complications [38]. In addition, the relationship between depression and lipid metabolism is further complicated by sex differences. A study focusing on first-diagnosed, drug-naïve depression patients revealed that female patients exhibited more pronounced changes in serum lipids during antidepressant therapy, with increased triglycerides and atherogenic index. This suggests that female patients with depression might be more susceptible to dyslipidemia, highlighting the importance of considering sex-specific factors in the management of depression-related metabolic disorders [39]. These findings collectively emphasize the critical need for integrated care approaches that address both mental health and metabolic health in young MDD patients. By recognizing and managing ALM in this population, healthcare providers can potentially mitigate the risk of long-term cardiovascular and metabolic complications, improving overall patient outcomes. Correlation between psychotic symptoms and ALM in young MDD patients In the present study, the prevalence of psychotic symptoms in MDD patients with combined ALM is nearly four times that of MDD patients without ALM. Research indicates that the prevalence of psychotic symptoms in MDD patients with ALM is notably elevated. For instance, a study examining the prevalence and clinical correlates of psychotic symptoms in first-episode and drug-naïve MDD patients found that those with comorbid ALM had a significantly higher incidence of psychotic symptoms compared to those without ALM [40]. This suggests a potential link between lipid metabolism abnormalities and the manifestation of psychotic features in MDD. Moreover, the presence of psychotic symptoms in MDD has been associated with various metabolic disturbances. A cross-sectional study involving young, first-episode, and drug-naïve MDD patients revealed that those with psychotic symptoms had higher levels of thyroid-stimulating hormone (TSH) and lipid abnormalities, such as elevated triglycerides and cholesterol levels, as compared to nonpsychotic MDD patients [28]. These findings underscore the complex interplay between metabolic dysfunction and psychiatric symptoms in MDD. In addition, the association between ALM and psychotic symptoms in MDD may be influenced by other factors such as insulin resistance. The triglyceride glucose (TyG) index, a marker of insulin resistance, has been linked to an increased likelihood of psychotic symptoms in MDD patients [41]. This further supports the hypothesis that metabolic dysregulation may contribute to the pathophysiology of psychotic symptoms in MDD. Our finding that ALM is associated with a nearly fourfold increase in psychotic symptoms (9.65% vs. 2.48%, P = 0.0069) resonates with a broader framework linking lipid dysregulation not just to mood, but to cognitive integrity. Psychotic symptoms can be conceptualized as a severe form of cognitive and perceptual disturbance. The work by Mehdi et al. [13] provides a compelling explanatory model for our observations. They found that depression itself can mediate the relationship between lipids and cognition, specifically showing that the neuroprotective effects of HDL on cognitive performance are absent in depressed individuals. It is plausible that in our cohort of young MDD patients, the underlying lipid dyshomeostasis contributes to a state of compromised neural health, and the concurrent depressive state negates any potential compensatory or protective lipid-related mechanisms, thereby lowering the threshold for the emergence of severe cognitive disturbances like psychosis. Furthermore, our identification of depression severity (HAMD scores) as a key risk factor for psychosis is indirectly supported by findings from Wagner et al. [12]. They demonstrated that higher LDL-C levels are associated with greater depression severity. This suggests a potential cascade: abnormal lipid profiles (e.g., high LDL) may contribute to a more severe depressive episode, which in turn, as our results show, significantly increases the risk of developing psychotic features. This positions lipid metabolism not merely as a comorbidity, but as a potential upstream factor influencing the clinical phenotype and severity of MDD. Comparison of findings between patients with and without abnormal lipid metabolism A notable advantage of this study is the comparison of MDD patients with and without ALM, which allowed us to identify both shared and distinct risk profiles for psychosis. A commonality across both groups was the strong association between psychotic symptoms and greater illness severity, evidenced by significantly higher HAMD and HAMA scores and a higher proportion of suicide attempts. This suggests that the severity of the depressive and anxiety components is a robust driver of psychosis, irrespective of metabolic status. However, a critical distinction emerged regarding endocrine factors. Elevated TSH was a significant and independent risk factor for psychosis only in the ALM group, a finding consistent with literature suggesting an intricate link between thyroid dysfunction, lipid metabolism, and the pathophysiology of severe psychiatric symptoms [42, 43]. This suggests the HPT axis’s role in psychosis may be potentiated by underlying metabolic dysregulation. It is important to note that a multivariate analysis for the non-ALM group was not performed. With only six patients exhibiting psychotic symptoms in this cohort, such an analysis would be statistically inappropriate due to an insufficient number of events per variable, which could lead to unreliable and potentially misleading results. Factors associated with psychotic symptoms in young MDD patients with ALM This study emphasizes the role of the thyroid-stimulating hormone (TSH) as a risk factor for psychotic symptoms in young major depressive disorder (MDD) patients, consistent with prior research on biological and clinical factors related to psychotic symptoms in MDD. Elevated TSH levels were significantly associated with psychotic symptoms in first-episode, drug-naïve MDD patients [44], and this association between thyroid function and psychiatric symptoms extends beyond MDD. For example, in young adults with anxious depression, TSH levels mediated the link between depression and psychotic symptoms [45]. In addition, the relationship between TSH and psychotic symptoms has been observed across different demographic groups and clinical settings, such as being a significant predictor in female young MDD patients [46]. These findings highlight the importance of TSH as a risk factor and suggest the need for further research on the mechanisms and therapeutic implications of thyroid dysfunction. Integrating thyroid function assessments into the clinical evaluation of MDD patients may improve identification and management of those at risk for psychotic symptoms [47, 48]. In the present study, the Hamilton Anxiety Rating Scale (HAMA) and Hamilton Depression Rating Scale (HAMD) scores were identified as significant risk factors for the development of psychotic symptoms in young patients with major depressive disorder (MDD) who also exhibit abnormal lipid metabolism (ALM). This finding is consistent with previous research that has explored the complex interplay between metabolic disturbances and psychiatric symptoms in depressive disorders. A study investigating the prevalence and risk factors of psychotic symptoms in first-episode, drug-naïve MDD patients found that higher HAMD and HAMA scores were significantly associated with the presence of psychotic symptoms [25]. This suggests that both anxiety and depression severity are critical components in the manifestation of psychotic features in MDD. For instance, a large-scale cross-sectional study on young, first-episode, drug-naïve outpatients with MDD demonstrated that psychotic symptoms were significantly associated with higher scores on both the HAMA and HAMD scales [48]. This study emphasized the importance of regular assessment of psychotic symptoms in young MDD patients, particularly those with elevated total cholesterol levels, as these factors were found to be interrelated. In conclusion, the findings from the present study are consistent with existing literature that highlights the significant role of HAMA and HAMD scores as risk factors for psychotic symptoms in young MDD patients with abnormal lipid metabolism. This emphasizes the need for comprehensive assessments that consider both psychiatric and metabolic factors in the management of MDD [49–51]. Limitation There are several limitations to the research. First, this was a cross-sectional study, so a causal relationship between psychotic symptoms and abnormal lipid metabolisms in patients with MDD cannot be established. Second, to minimize the influence of other factors on lipids, we excluded subjects with severe physical illness, drug or alcohol abuse histories, as well as pregnant or breastfeeding women, who might have impacted the results. This does not, however, rule out the possibility that other confounding factors may have influenced our findings, particularly factors affecting lipid metabolism, such as smoking and alcohol consumption, social status, and household income, which have a significant impact on lipid levels, psychiatric symptoms, and depression. It is anticipated that future studies will include these patients and conduct the necessary follow-up studies to examine the effect of lipids on psychiatric symptoms in adolescents with PMD. Third, the generalizability of our findings is constrained by the specific characteristics of our study sample. Our cohort was recruited from a single center and consisted exclusively of Han Chinese outpatients. This presents two main limitations. Primarily, focusing solely on the Han Chinese population means our results may not be readily applicable to other ethnic groups, given that genetic predispositions and environmental factors that influence both lipid metabolism and psychiatric disorders can vary significantly across populations. Secondly, by recruiting only outpatients, our sample may not be representative of the full spectrum of disease severity, particularly more acutely ill inpatients or individuals with milder symptoms in the general community. Therefore, caution is warranted when extrapolating our findings to non-Chinese populations or different clinical settings. Fourth, in this preliminary study, there was no control group, and future longitudinal studies will investigate the relationship between lipids and psychiatric depression in patients with PMD. Fifth, additionally to delusions and hallucinations, patients with PMD are more likely to exhibit specific features, such as pathological cognitive and psychomotor disorders; therefore, in the future, scale analysis or potential discriminatory clinical variables should be considered. Some studies have demonstrated differences in pathogenic mechanisms and pathophysiological changes between negative and positive psychotic symptoms; however, due to a small sample size, this study was unable to conduct a subgroup analysis of negative and positive psychotic symptoms. Conclusion In conclusion, our study demonstrates that young MDD patients with abnormal lipid metabolism exhibit a high prevalence of psychotic symptoms. The factors significantly associated with psychotic symptoms include HAMD, HAMA, and TSH levels. The combined use of these three factors yielded a higher AUC, highlighting their potential as reliable predictors of psychotic symptoms. These findings underscore the importance of these clinical factors in the prediction of psychotic symptomatology in young MDD patients. Future research is needed to explore the underlying mechanisms and validate these findings in larger, more diverse cohorts. Acknowledgements We are grateful to the doctors and nurses who contributed to the data collection in this study, as well as to all the participants who took part in the survey. Author contributions Boxuan Li: Writing—original draft, Formal analysis, Conceptualization. Minghui Li: Writing—original draft, Formal analysis, Conceptualization. Mengqian Li: Writing—original draft, Formal analysis, Conceptualization. Xing Wang: Data curation, Supervision, Formal analysis. Mingxiu Kang: Data curation, Investigation, Supervision. Guoshuang Zhang: Investigation, Methodology, Writing—review & editing. Guangming Xu: Writing—Review & editing, Data curation. Xiangyang Zhang: Writing—Review & editing, Data curation, Conceptualization. Funding This work was supported by the Tianjin Health Research Project (Grant No. TJWJ2024QN082), the Tianjin Municipal Education Commission Fund (Grant No. 2023KJ046), and the Tianjin Key Medical Discipline Construction Project (Grant No. TJYXZDXK-3-015B). These sources played no additional part in the study’s conception, data colleting and analysis, report writing, or decision to submit the work for publication. Availability of data and materials The data that support the findings of this study are available from the corresponding author upon reasonable request. Declarations Ethics approval and consent to participate The research was approved for clearance by the First Affiliated Hospital of Shanxi Medical University’s Institutional Review Board (No. Y27). Consent for publication Each patient signed a written informed consent form after being informed of the study procedures. Competing interests The authors declare no competing interests. Footnotes Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Boxuan Li, Minghui Li, and Mengqian Li have contributed equally to this work and share first authorship. Contributor Information Guangming Xu, Email: xugm@tmu.edu.cn. Xiangyang Zhang, Email: zhangxy@psych.ac.cn. References Associated Data Data Availability Statement The data that support the findings of this study are available from the corresponding author upon reasonable request.