Temporal trends of blood-based markers in various mental disorders and their relationship with brain structure

doi:10.21203/rs.3.rs-4321120/v1

Download PDF

Article

Temporal trends of blood-based markers in various mental disorders and their relationship with brain structure

https://doi.org/10.21203/rs.3.rs-4321120/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background:

Studies have identified blood-based biomarkers for various mental disorders, but their temporal trends and relationship with brain structure remain unclear. This study aimed to assess the temporal trends of blood-based biomarkers across 10 years leading up to and following diagnosis and explore their association with brain structure.

Methods:

We conducted a nested case-control study using prospective community-based cohort data from UK Biobank (n = 502,617; aged 40 to 69 years; recruited from 2006 to 2010), which included both psychiatric assessments and blood-based biomarkers. Cases were defined as individuals with a diagnosis of mental disorders at baseline and during follow-up (individuals with bipolar disorder = 1,325; depression = 36,582; schizophrenia = 1,479; anxiety = 27,220). Nearly 5 controls without any mental disorders were matched for each case. Multivariable linear regression was used to assess the divergence evolution between cases and controls for each psychiatric assessment and blood-based biomarker.

Results:

In comparison to controls, 6, 15, 10, and 47 blood-based markers exhibited significant changes over time in bipolar disorder, anxiety, schizophrenia, and depression, respectively. These biomarkers could be grouped into distinct clusters with complex, non-linear temporal trends. Some clusters displayed monotonic changes, while others reversed near the time of diagnosis. The identified blood-based markers were associated with brain structure in the general population, including orbitofrontal, precuneus, and amygdala regions.

Conclusions:

These findings provide novel insights into the temporal trends of blood-based biomarkers in various mental disorders within 10 years before and after clinical diagnosis, as well as their correlations with brain structure. Monitoring and managing these biomarkers could potentially carry significant implications for the early detection and prevention of mental disorders in older adults.

More than 14% of adults aged 60 and above suffer from a mental disorder, and the prevalence increases with age^{[1, 2]}. Considering the rapid aging of the global population^[3], it is expected that a growing number of elderly individuals will experience mental disorders in the coming years. Mental disorders are often accompanied by a multitude of debilitating symptoms^[4], including emotional dissonance, cognition decline, motor dysfunction and disturbances of the metabolic and immune system^[5–10]. These alterations not only have an impact on health in their own right but also increase the risk of other adverse health outcomes, encompassing neurodegenerative disorders^[11], cardiovascular disease^{[12, 13]} and cognitive impairment^[2]. They can also result in a substantial reduction in life expectancy (10–20 years)^[14–16]. Since mental disorders are closely linked to poor health outcomes, it is crucial to comprehend the contributing factors in order to advance the prevention and treatment of these disorders.

Older adults frequently face underrecognition and undertreatment of their mental health conditions, yet early intervention at the early signs of mental disorders can improve a range of outcomes^[16]. Recent evidence has revealed changes in metabolic markers in mental disorders compared with controls and differ significantly between different mental disorders ^[17–20]. Furthermore, previous studies have also identified causal roles for certain metabolites in mental disorders, suggesting that metabolic dysregulation contribute to the development of mental disorders^{[10, 21, 22]}. Various immune biomarkers such as C-reactive protein, immune cell count have also been associated with mental disorders^[23–26]. The intricate interaction between metabolism and immunity adds further complexity to this field^{[27, 28]}. Previous research has documented the pre-diagnosis trajectories of select blood markers in depression, anxiety, and stress-related disorders^[19]. However, a comprehensive understanding of the temporal trends of blood based markers throughout the pre- and post-diagnostic phases of various mental disorders is currently lacking. Additionally, the association between distinct blood markers and brain imaging findings remains uninvestigated. Understanding the temporal trends of these markers holds promise for enhancing the early diagnosis and may ultimately inform tailored treatment strategies in mental disorders.

The present study aims to explore the temporal trends of psychiatric assessments and various blood-based markers, encompassing blood cell count, blood biochemistry, and metabolomic markers, among individuals suffering from various mental disorders. By utilizing data from the UK Biobank, encompassing over 500,000 elderly adults, including those diagnosed with various mental disorders, we performed a nested case-control analysis. This approach enabled us to assess the temporal trends of blood-based markers and identify potential patterns unique to specific mental disorders. Furthermore, we investigated the relationship between indentified blood-based markers and brain structure using neuroimaging data.

Study cohort

Participants of this study were extracted from the UK Biobank (UKB), a large prospective cohort study consisting of 502,617 participants from 22 assessment centers aged between 40 and 69 years at baseline from March 2006 to October 2010^[29]. The dataset encompasses comprehensive information on participants' demographic, behavioral, blood, neuroimaging assessments, and health-related outcomes. To ensure the reliability of the study, individuals with mental disorders who did not have a follow-up record were excluded. Additionally, to avoid potential sample bias, individuals who were diagnosed with mental disorders more than 10 years before or after the baseline visit were also excluded. The flow chart of eligible standards was demonstrated in Supplementary Fig. 1. UKB received ethical approval from the National Health Service North West Multi-Center Research Ethics Committee and the National Information Governance Board for Health and Social Care. All participants provided written informed consent, emphasizing the voluntary nature of participation and the freedom to withdraw at any time without explanation.

Outcome ascertainment

We explored four mental disorders, including anxiety, depression disorders, bipolar disorders and schizophrenia. Diagnosis and recorded dates were ascertained and classified according to the ICD-10 codes (Table S1), extracted from first occurrences data (UKB category 2401–2417) that mapped from data sources of primary care (category 3000), hospital inpatient (category 2000), self-reported medical conditions (UKB field 20002) and death register records (field 40001 and 40002). The 10-level subchapter categories of mental disorders were described as codes F10-19, F20, F30-F39, etc. in ICD-10. The date of any earliest recorded diagnosis listed above was leveraged as a proxy for the actual date of diagnosis.

Blood-based biomarker measurements

Blood samples of participants were collected at baseline (2006–2010) by the UK Biobank. The processing details of blood samples can be found on the UKB website (https://www.ukbiobank.ac.uk/enable-your-research/about-our-data/biomarker-data). In this study, 31 blood cell counts, 29 biochemistry markers, and 168 serum metabolites commonly were used. Standard hematological assays, such as full blood counts and related parameters, were performed on fresh whole blood within 24 hours of blood collection for all participants. Multiple immunoassays and clinical chemistry analyzers were used to measure blood biochemistry markers, details of which are provided in the UKB website (https://www.ukbiobank.ac.uk/enable-your-research/about-our-data/biomarker-data). The high-throughput targeted nuclear magnetic resonance (NMR) metabolomics platform (Nightingale Health Ltd) was used to generate spectra from which 168 metabolic measures were simultaneously quantified in a single assay, either as absolute concentrations of each metabolic measure or as ratios. These quantitative metabolites include lipids, lipoprotein, fatty acids, apolipoproteins, and various low-molecular-weight metabolites such as amino acids, glycolysis-related metabolites, and ketone bodies. Metabolites were measured in a random sequence according to standard biochemical analysis protocols, and laboratory staff were blinded to the related health outcomes. The FieldID of metabolites and serum assessments were displayed in Supplementary Table 2, while abbreviations of all Blood-based Markers were demonstrated in Supplementary Table 2.

Psychiatric assessments

The Patient Health Questionnaire-4^[30] (PHQ-4) and the Eysenck Personality Questionnaire-Revised (EPQ-R)^[31] were utilized to evaluate self-reported symptoms of psychiatric traits. The PHQ-4 instrument, a 4-item inventory, comprises a 2-item depression scale (PHQ-2) and a 2-item anxiety scale (general anxiety disorder [GAD]-2), rated on a 4-point Likert-type scale. The EPQ-R scale, designed to assess neuroticism, comprises 12 items. The specific items used to assess each symptom are provided in Supplementary Table 3.

Neuroimaging phenotypes.

The neuroimaging phenotypes encompassed two main categories: (1) regional grey matter volumes, which included 39 major structures; and (2) white matter microstructure, indexed by fractional anisotropy (FA), mean diffusivity (MD), intra-cellular volume fraction (ICVF), as measures of white matter microstructure. All neuroimaging data were derived from the UK Biobank Imaging Database.

The UK Biobank protocol for conducting brain MRI was employed. The detailed protocol is available at https://biobank.ctsu.ox.ac.uk/crystal/refer.cgi?id=2367. UK Biobank acquired, pre-processed, and quality controlled the images using FMRIB Software Library packages^[32] under a standard protocol. Only data from the pilot study that had consistent scanner settings and passed the initial quality assessment conducted by the UK Biobank imaging team were included in the analysis.

All imaging data were collected using a 3-T Siemens Skyra (software platform VD13) machine. In this analysis, individuals with conditions confounding the assessment of brain atrophy were excluded. These conditions included brain tumor, neurodegenerative disease, cerebrovascular diseases, cerebral infarction, epilepsy, multiple sclerosis, encephalitis or brain abscess, which were identified using ICD-10 codes and self-reported diseases (Supplementary Table 4).

Propensity score matching

For each participant with a mental disorder, nearly 5 controls were selected based on propensity score matching for age, sex, ethnicity, centers, years of education, body mass index (BMI), socioeconomical status, drinking status and smoking status using the R package “Matchit”^[33]. Controls were defined as individuals without blood &immune disorders, mental & behavioral disorders, disease of infections and nervous system disorders. We used the baseline date and the diagnosis date to calculate the years from clinical diagnosis (date of clinical diagnosis minus baseline date). The time since diagnosis of control participants were arranged as the time since diagnosis of their matched patients. Consequently, each patient and matched control participants had the same index date. This provided a dataset of matched case-control populations under consecutive time windows before and after diagnosis, enabling further investigation of divergence evolution over time between cases and controls for each psychiatric assessment and blood-based biomarker (Supplementary Fig. 2).

Statistical analysis

Our primary objective was to identify measures of divergence evolution between individuals with mental disorders and controls. We used a multivariable linear regression model to compare each variable of interest between the two groups. We included group and its interactions with time and the square of time as independent variables, with blood-based markers as the response variable. Other independent variables included age, sex, ethnicity, years of education, Townsend deprivation index, study site, BMI, drinking status, and smoking status at the baseline visit. Ethnicity was categorized as white, asian, black, and other ethnicity (e.g., mixed ethnicity or other ethnicity). Continuous variables were normalized before regression, and categorical variables were made dummy variables. Missing values were not included in the analysis. P-values for group and its interactions with time and the square of time were corrected for multiple comparisons using the Bonferroni method, respectively. Blood-based markers with any corrected p-values less than 0.01 were used for temporal trends visualization. The analysis was performed using the StatsModel (v0.11.1) and ScikitLearn (v0.24.1) packages in Python (v3.9).

To visualize the temporal trends of blood-based markers in mental disorders, we employed the locally estimated scatterplot smoothing (LOESS) method to assess linear or non-linear patterns in blood-based marker evolution over time. To ensure that each assessment's values were easily comparable, blood-based marker values for mental disorders were transformed using z-scores relative to controls, adjusting for the aforementioned covariates. To reduce the complexity of blood-based markers, we grouped blood-based markers with similar temporal trends across diseases using unsupervised hierarchical clustering via the hclust function from the R stats package. The correlation between the identified blood-based markers and brain structures were performed using Pearson's correlation adjusted for the aforementioned covariates.

Role of the funding source

The funder of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report.

Study populations

In this study, 56,553 individuals diagnosed with mental disorders were enrolled, with a median age of 57.00 [IQR 49.00–62.00]. Among them, 20,533 (36.31%) were male, and the majority (95.32%) had white ethnicity. The sample included 27,220 individuals diagnosed with anxiety, 1,325 with bipolar disorder, 36,582 with depression, and 1,479 with schizophrenia. Table 1 presents the demographic information. Nearly 5 controls were matched for each case. Therefore, a total of 255,867 individuals were included in the main analyses. Similar demographic statistics were observed between the case and control groups (Table 1).

Table 1

The characteristics of UK-Biobank participants with mental disorders and controls.
	Anxiety		Bipolar
	Anxiety subjects (n = 27,220)	Controls (n = 131,941)	Bipolar subjects (n = 1,325)	Controls (n = 6,559)
Age at baseline (years)	57.0 [49.0, 63.0]	57.0 [50.0, 63.0]	55.0 [48.0, 62.0]	55.0 [48.0, 62.0]
Gender (male, %)	9,220 [33.9%]	45,193 [34.3%]	562 [42.4%]	2,780 [42.4%]
Townsend deprivation index	-1.7 [-3.4, 1.3]	-1.9 [-3.5, 1.0]	-0.6 [-2.9, 2.7]	-0.9 [-2.9, 2.4]
BMI	27.1 [24.3, 30.6]	26.9 [24.3, 30.2]	27.8 [24.8, 31.6]	27.5 [24.8, 31.0]
Ethnicity
White	25,986 [95.5%]	126,383 [95.8%]	1,248 [94.2%]	6,212 [94.7%]
Asian	325 [1.2%]	1,482 [1.1%]	18 [1.4%]	77 [1.2%]
Black	518 [1.9%]	2,416 [1.8%]	21 [1.6%]	101 [1.5%]
Others	391 [1.4%]	1,660 [1.3%]	38 [2.7%]	169 [2.6%]
Education (years)
< 10	5,754 [21.1%]	25,014 [19.0%]	224 [16.9%]	1,039 [15.8%]
10–15	12,047 [44.3%]	59,956 [45.4%]	527 [39.8%]	2,654 [40.5%]
> 15	9,419 [34.6%]	46,971 [35.6%]	574 [43.3%]	2,866 [43.7%]
Number of individuals (years before and after diagnosis)
-10~-5	9,658 [35.5%]	46,746 [35.4%]	354 [26.7%]	1,760 [26.8%]
-5 ~ 0	6,424 [23.6%]	31,194 [23.6%]	328 [24.8%]	1,623 [24.7%]
0 ~ 5	6,377 [23.4%]	30,906 [23.4%]	349 [26.3%]	1,718 [26.2%]
5 ~ 10	4,761 [17.5%]	23,095 [17.5%]	294 [22.2%]	1,458 [22.3%]
	Depression		Schizophrenia
	Depression subjects (n = 36,582)	Controls (n = 168,985)	Schizophrenia subjects (n = 1,479)	Controls (n = 7,276)
Age at baseline (years)	56.0 [49.0, 62.0]	56.0 [49.0, 62.0]	56.0 [48.0, 62.0]	56.0 [48.0, 63.0]
Gender (male, %)	13,649 [37.3%]	64,678 [38.3%]	753 [50.9%]	3,683 [50.6%]
Townsend deprivation index	-1.5 [-3.3, 1.7]	-1.9 [-3.4, 1.0]	1.0 [-2.3,4.1]	0.5 [-2.5, 3.5]
BMI	27.5 [24.7, 31.2]	27.1 [24.5, 30.4]	27.8 [24.7, 31.3]	27.6 [24.8, 30.8]
Ethnicity
White	34,907 [95.4%]	161,820 [95.8%]	1,310 [88.6%]	6,494 [89.2%]
Asian	418 [1.1%]	1,869 [1.1%]	70 [4.7%]	333 [4.6%]
Black	635 [1.7%]	2,907 [1.7%]	44 [3.0%]	200 [2.8%]
Others	622 [1.7%]	2,389 [1.4%]	55 [3.7%]	249 [3.4%]
Education (years)
< 10	7,734 [21.1%]	30,396 [18.0%]	373 [25.2%]	1,562 [21.5%]
10–15	15,950 [43.6%]	75,843 [44.9%]	568 [38.4%]	3,065 [42.1%]
> 15	12,898 [35.3%]	62,746 [37.1%]	538 [36.4%]	2,649 [36.4%]
Number of individuals (years preceding diagnosis)
-10~-5	8,397 [23.0%]	39,123 [23.2%]	527 [35.6%]	2,608 [35.8%]
-5 ~ 0	6,679 [18.3%]	30,899 [18.3%]	374 [25.3%]	1,834 [25.2%]
0 ~ 5	12,277 [33.6%]	56,578 [33.5%]	275 [18.6%]	1,364 [18.7%]
5 ~ 10	9,229 [25.2%]	42,385 [25.1%]	303 [20.5%]	1,470 [20.2%]

Temporal trends of psychiatric assessments in mental disorders before and after clinical diagnosis

The results of two psychiatric assessments demonstrate significant divergent evolution over time between individuals with mental disorders and controls. Among the four mental disorders, neuroticism scores continued to increase in the 10 years before diagnosis (Fig. 1). In patients with anxiety and depression, neuroticism scores continued to increase after diagnosis, but the rate of increase slowed down. In patients with schizophrenia and bipolar disorder, neuroticism scores began to decrease around the time of diagnosis, although the decrease was not significant. The temporal trend of PHQ-4 follows an inverted U-shape with different disorders showing that reversed around the time of clinical diagnosis (Fig. 1).

Temporal trends of blood-based makers in mental disorders before and after clinical diagnosis

The evolutionary temporal trends of 6, 15, 10 and 47 blood-based markers identified for bipolar disorder, anxiety, schizophrenia, and depression were distinct from those of the control group. These blood-based marker were grouped into 4 to 6 clusters for different mental disorders according to their temporal trends, with sizes ranging from 1 to 15 blood-based markers (Fig. 2). Although we employed a standard linear model to identify these blood-based markers, nearly all clusters displayed non-linear temporal trends within the 10-year window before and after clinical diagnosis.

In patients with anxiety disorders, the temporal trends of cluster 1 displayed a general downward trend over time, with a reduction in the rate of slowed down after diagnosis, including measures such as white blood cell count (WBC), neutrophill percentage (NEUT-P), mean sphered cell volume (MSCV), mean reticulocyte volume (MRV), free cholesterol in small ldl (S-LDL-FC), and Urate. The temporal trends of clusters 2 and 4 revealed a U-shaped evolution that reversed around the time of clinical diagnosis. Cluster 2 encompassed mean corpuscular volume (MCV) and glycated haemoglobin (HbA1c), while cluster 4 included Triglycerides, mean corpuscular haemoglobin (MCH), aspartate amino transferase (AST), and gamma glutamyl transferase (GGT). Conversely, the temporal trends of clusters 3 presented an inverted U-shape, including markers such as immature reticulocyte fraction (IRF), IGF-1, red blood cell count (RBC), and Vitamin D. Additionally, the temporal trends of cluster 5 incorporated lymphocyte percentage (LYMP-P) and high light scatter reticulocyte count (HLR-C), which increased in the 10 years before diagnosis and did not show further increase after diagnosis. Finally, in cluster 6, the temporal trends of haemoglobin concentration (Hb) demonstrated undulating changes, while the temporal trend of free cholesterol in medium ldl (M_LDL_FC) presented an inverted U-shape.

In patients with depression, the temporal trends of cluster 1 and 5 followed an inverted U-shape, with levels peaking around 2 years after diagnosis before decreasing. This cluster included biomarkers like Cholesterol, ldl cholesterol (LDL-C), and triglycerides in ldl (IDL-TG). Additionally, cluster 5 incorporated biomarkers like Acetate, red blood cell distribution width (RDW), mean corpuscular volume (MCV), and so on, whose temporal trends also presented an inverted U-shape with levels increasing prior to diagnosis but decreasing afterwards. However, compared to cluster 1, this cluster exhibited a lower rate of change. The temporal trends of cluster 2 exhibited a general upward trend over time, with a particularly notable increase in levels prior to diagnosis. This cluster included biomarkers such as total protein, alkaline phosphatase (ALP), aspartate aminotransferase (AST), mean corpuscular haemoglobin concentration (MHCH), and lymphocyte percentage (LYMP-P). On the other hand, the temporal trends of cluster 3 displayed a downward trend, encompassing biomarkers like urate, estosterone, c-reactive protein (CPR), high light scatter reticulocyte count (HLR-C), and immature reticulocyte fraction (IRF). Finally, the temporal trends of cluster 4 and 6 continued to decrease with a fast rate before diagnosis and then increased slightly. Cluster 4 incorporated blood-based markers like NEUT-P, SHBG, WBC, and vitamin D, while cluster 6 consisted of markers such as total bilirubin, mean platelet volume (MPV), Hb, and RBC.

In patients with schizophrenia, the temporal trends of cluster 1 followed an inverted U-shape, with levels peaking around 2 years after diagnosis before decreasing. This cluster included blood markers such as Acetate and red blood cell distribution width (RDW). Conversely, the temporal trends of cluster 2 followed a U-shape, with the lowest level between 0 and 2 years after diagnosis. This cluster included markers such as Hb, haematocrit percentage (HCT), WBC, and cysytain C. The temporal trends of cluster 3 increased before diagnosis and then decreased 2–4 years after diagnosis, including measures such as ALP, CPR, MSCV, and MRV. On the other hand, the temporal trends of cluster 4 demonstrated an inverted U-shape over time between − 10 and 4 years, including factors such as HbA1c and GGT. Finally, the temporal trends of Cluster 5 exhibited a downward trend, encompassing only Vitamin D.

In patients with bipolar disorder, the temporal trends of biomarkers can be categorized into four distinct clusters. Cluster 1 comprises markers such as RBC, Hb, Hct, and IRF, which follow a U-shaped temporal trends with the lowest values occurring around the diagnosis date. Cluster 2 incorporates markers of MCV and MSCV, whose temporal trends exhibit a downward trend. On the other hand, cluster 3 comprises a single blood marker, high light scatter reticulocyte percentage (HLR-P), which follows an upward temporal trends.

Identified blood-based makers were associated with brain structures

We found significant positive relationships between markers linked to lipid metabolism—specifically, cholesterol and LDL-C levels—and the volume of numerous cortical brain regions. Additionally, immune markers such as neutrophill count (NEUT-C), lymphocyte count (LYMP-C), and CPR were found to be correlated with the volume of several brain regions, including the orbitofrontal cortex and hippocampus. Notably, the level of vitamin D in the blood was positively correlated with the volume of the medial orbitofrontal cortex. Furthermore, an elevation in WBC was associated with a reduction in the volume of orbitofrontal and hippocampus regions, as well as an increase in ventricular volume (Fig. 3A).

Additionally, we found significant relationships between the MD values of the uncinate fasciculus (UF), superior thalamic radiation (STR), and superior longitudinal fasciculus (SLF) and most blood-based markers (Fig. 3). The ICVF values of the parahippocampal part of cingulum (PPOC), middle cerebellar peduncle (MCP), and corticospinal tract were associated with certain blood cell counts, including red and white blood cell counts, hemoglobin concentration, haematocrit percentage, NEUT-C and LYMP-C. Furthermore, liver function markers such as ALP and ALT, as well as vitamin D levels, were also linked to these brain structures. Notably, the FA values of the MCP and forceps major were found to be associated with lipid metabolism indicators, immune markers, and liver function markers, including triglycerides, CRP, and GGT (Fig. 3B).

In this study, we investigated temporal trends of psychiatric assessments and blood-based markers spanning a 10-year period before and after clinical diagnosis of mental disorders. Notably, the majority of these markers demonstrated nonlinear patterns of change, with a subset exhibiting a distinct U-shaped temporal trend that culminated around the time of clinical diagnosis. Among the blood-based markers identified, lipid metabolic markers emerged as the most strongly correlated with brain structures, followed by markers related to immune function. The results suggested that most biomarkers begin to change years before diagnosis, possibly because there is often a delay of years between the onset and diagnosis of mental disorders or a maladaptive response^[19].

Neuroticism has been suggested to serve as an efficient marker of non-specified general risk for mental disorders^[34]. We observed an upward trend of psychiatric assessments in four different mental disorders prior to diagnosis, indicating a worsening of symptoms as the disease progressed. For patients with anxiety and depression disorders, the neuroticism score continued to increase, but the rate of increase slowed down after diagnosis. However, following diagnosis, the PHQ-4 displayed a downward trend across four mental disorders. In line with temporal trends of psychiatric assessments, we observed changes in several blood-based markers that occurred up to a decade before diagnosis, presenting an inverted or positive U-shaped trend. This finding raises the possibility that these changes alleviate after treatment, or changes of these markers may be compensatory responses to developing illness^[35–38].

The findings from our study reveal a significant downward trend in Hb, Hct, and RBC in patients with depression, schizophrenia, and bipolar disorder prior to diagnosis. However, these parameters gradually improved following diagnosis. In patients with anxiety disorders, a similar trend was observed for MCV, HbA1c, and MCH, which also improved post-diagnosis. These results suggest that patients with mental disorders may have some degree of anemia and iron deficiency prior to their diagnosis. This is consistent with previous research that has linked iron deficiency or anemia with an increased risk of developing mental disorders^[39–42]. We found positive correlations between RBC, Hb, Hct, and the ICVF values of ML, MCP, and corticospinal tract. These findings suggest that a decrease in these blood parameters may be associated with a reduction in neurite density in these tracts. Additionally, our also found a positive correlation between RBC and the volume of the amygdala, suggesting that a decrease in the number of red blood cells may contribute to the atrophy of this key brain region involved in the regulation of emotions^{[43, 44]}. Therefore, monitoring and intervention on blood oxygen levels in individuals with high risk of mental disorders may help prevent further deterioration.

Two immune cells, LYMP-P and NEUT-P, that were significantly altered in anxiety and depression patients compared to controls. This finding is consistent with previous research that suggests inflammation is associated with an increased risk of depression and anxiety^[24]. Specifically, LYMP-P and NEUT-P were found to have a significant correlation with the volume of the medial and lateral orbitofrontal cortex, key brain regions involved in the representation of reward value and non-reward^[45]. Reduced orbitofrontal volumes and functional impairment have been observed in patients with mood disorders^{[46, 47]}. Furthermore, our analysis revealed significant correlations between these two immune cells and the volume of the hippocampus. Our previous studies found that both depression and schizophrenia have a subtype of brain atrophy that starts from the hippocampus^{[48, 49]}. Considering emerging evidence indicating that inflammation can contribute to the pathogenesis of depression and that drugs with primary immune targets may prevent or improve depressive or anxiety symptoms^{[24, 50]}.

In previous studies, inadequate vitamin D levels have been associated with an increased risk of various mental disorders^{[51, 52]}. We found the level of vitamin D in the blood of patients with schizophrenia and depression continued to decrease before diagnosis. Furthermore, there was a positive correlation between the level of vitamin D and the volume of the medial orbitofrontal cortex, as well as a negative correlation with the ICVF value of PPOC, MCP, and forceps major. Previous studies have also found that vitamin D supplementation can improve total attention span and positive and negative symptoms in schizophrenia^[53]. Therefore, individuals at high risk of schizophrenia and depression should be encouraged to obtain sunlight exposure to maintain optimal vitamin D levels.

Strengths and Limitations

This study presents several strengths, including the utilization of a community-based cohort with a substantial sample size, a long follow-up period, and the comprehensive evaluation of blood-based markers. However, the results should be interpreted within the context of several limitations. Firstly, as we used a cross-sectional dataset to explore temporal trends of blood-based markers, the ability to detect temporal trends may be limited, especially with fewer observations at certain time ranges. Additionally, the diagnosis time of mental disorders does not always align with the onset of disorders, potentially affecting the accuracy of blood-based marker trends. Future studies with follow-up blood samples are crucial to validate the temporal trends of these markers. Secondly, the UKB participants were predominantly white, and the psychiatric assessments were relatively simple. Future prospective studies with more diverse populations and comprehensive psychiatric assessments are required to further validate their temporal trends. Additionally, daily biological fluctuations in blood-based markers may limit the robustness of temporal trends analysis. Future studies should incorporate biological fluctuations of blood-based markers with repeated measurement of data. Finally, it's worth noting that the observed changes in blood-based markers may be influenced by various confounding factors, such as lifestyle habits, comorbidities, and medication use. Controlling for these factors in future studies would help to clarify the specific role of mental disorders in altering blood-based markers.

In conclusion, our study provides insights into the complex temporal trends of blood-based markers in individuals with mental disorders and their association with brain structure. Consistent monitoring and proactive management of these biomarkers could potentially facilitate early intervention and prevent further deterioration.

Contributors

LBW and WC conceived, designed, and supervised the project. YJW and JY performed statistical analyses. LBW and YJW drafted the manuscript. JY, JJK and YJW accessed and verified the underlying data reported in the manuscript. All authors had full access to all the study data and accepted the responsibility to submit it for publication.

Declaration of interests

The authors declare no conflict of interest related to this work.

Role of the funding source

The funder had no role in study design, data collection, data analysis, data interpretation, writing of the report, and decision to submit the paper for publication.

Acknowledgments

This study was funded by grants from the National Natural Science Foundation of China (82302312 and 82071997), the National Key R&D Program of China (2019YFA0709502), the Natural Science Foundation of Shanghai (23ZR1406000), Shanghai Municipal Science and Technology Major Project (2018SHZDZX01), Shanghai Rising-Star Program (21QA1408700) and 111 Project (B18015). Further, we would like to thank the support from the Shanghai Center for Brain Science and Brain-Inspired Technology, ZHANGJIANG LAB, and the State Key Laboratory of Neurobiology and Frontiers Center for Brain Science of Ministry of Education. This study utilized the UK Biobank Resource under application number 19542. We want to thank all the participants and researchers from the UK Biobank. The language of this paper has been polished by ERNIE Bot (v3.5) to enhance its readability and overall linguistic quality. The authors are responsible and accountable for the originality, accuracy, and integrity of the work.

Data sharing

All data used in this study were accessed from the publicly available UK Biobank Resource under application number 19542. These data cannot be shared with other investigators.

Institute of Health Metrics and Evaluation. Global Health Data Exchange (GHDx). https://vizhub.healthdata.org/gbd-results/, accessed 20 October 2023).
Kulak-Bejda, A., G. Bejda, and N. Waszkiewicz, Mental Disorders, Cognitive Impairment and the Risk of Suicide in Older Adults. Front Psychiatry, 2021. 12: p. 695286.
Jarzebski, M.P., et al., Ageing and population shrinking: implications for sustainability in the urban century. npj Urban Sustainability, 2021. 1(1).
Mane-Damas, M., et al., Autoimmunity in psychotic disorders. Where we stand, challenges and opportunities. Autoimmun Rev, 2019. 18(9): p. 102348.
Sheppes, G., G. Suri, and J.J. Gross, Emotion regulation and psychopathology. Annu Rev Clin Psychol, 2015. 11: p. 379-405.
Hirjak, D., et al., Motor dysfunction as research domain across bipolar, obsessive-compulsive and neurodevelopmental disorders. Neurosci Biobehav Rev, 2018. 95: p. 315-335.
Fett, A.J., et al., Long-term Changes in Cognitive Functioning in Individuals With Psychotic Disorders: Findings From the Suffolk County Mental Health Project. JAMA Psychiatry, 2020. 77(4): p. 387-396.
Karshikoff, B., M. Lekander, and J. Lasselin, Editorial: Vulnerability and protective factors for inflammation-associated somatoform and mental disorders. Brain Behav Immun, 2022. 106: p. 227-230.
Chourpiliadis, C., et al., Metabolic Profile and Long-Term Risk of Depression, Anxiety, and Stress-Related Disorders. JAMA Netw Open, 2024. 7(4): p. e244525.
Davyson, E., et al., Metabolomic Investigation of Major Depressive Disorder Identifies a Potentially Causal Association With Polyunsaturated Fatty Acids. Biol Psychiatry, 2023. 94(8): p. 630-639.
Saiz-Vazquez, O., et al., Depression as a Risk Factor for Alzheimer's Disease: A Systematic Review of Longitudinal Meta-Analyses. J Clin Med, 2021. 10(9).
Nielsen, R.E., J. Banner, and S.E. Jensen, Cardiovascular disease in patients with severe mental illness. Nat Rev Cardiol, 2021. 18(2): p. 136-145.
Kwapong, Y.A., et al., Association of Depression and Poor Mental Health With Cardiovascular Disease and Suboptimal Cardiovascular Health Among Young Adults in the United States. J Am Heart Assoc, 2023. 12(3): p. e028332.
Plana-Ripoll, O., et al., A comprehensive analysis of mortality-related health metrics associated with mental disorders: a nationwide, register-based cohort study. Lancet, 2019. 394(10211): p. 1827-1835.
Wykes, T., et al., Mental health research priorities for Europe. Lancet Psychiatry, 2015. 2(11): p. 1036-42.
Zhang, Z., et al., Depressive Symptoms and Mortality Among US Adults. JAMA Netw Open, 2023. 6(10): p. e2337011.
Amin, N., et al., Interplay of Metabolome and Gut Microbiome in Individuals With Major Depressive Disorder vs Control Individuals. JAMA Psychiatry, 2023. 80(6): p. 597-609.
Nedic Erjavec, G., et al., Short overview on metabolomic approach and redox changes in psychiatric disorders. Redox Biol, 2018. 14: p. 178-186.
Bot, M., et al., Metabolomics Profile in Depression: A Pooled Analysis of 230 Metabolic Markers in 5283 Cases With Depression and 10,145 Controls. Biol Psychiatry, 2020. 87(5): p. 409-418.
Tomasik, J., et al., Metabolomic Biomarker Signatures for Bipolar and Unipolar Depression. JAMA Psychiatry, 2023.
Xiao, G., et al., Causality of genetically determined metabolites on anxiety disorders: a two-sample Mendelian randomization study. J Transl Med, 2022. 20(1): p. 475.
Marijnissen, R.M., et al., Metabolic dysregulation and late-life depression: a prospective study. Psychol Med, 2017. 47(6): p. 1041-1052.
Wittenberg, G.M., et al., Major Depressive Disorder Is Associated With Differential Expression of Innate Immune and Neutrophil-Related Gene Networks in Peripheral Blood: A Quantitative Review of Whole-Genome Transcriptional Data From Case-Control Studies. Biol Psychiatry, 2020. 88(8): p. 625-637.
Milaneschi, Y., et al., Association of inflammation with depression and anxiety: evidence for symptom-specificity and potential causality from UK Biobank and NESDA cohorts. Mol Psychiatry, 2021. 26(12): p. 7393-7402.
Haapakoski, R., et al., Cumulative meta-analysis of interleukins 6 and 1beta, tumour necrosis factor alpha and C-reactive protein in patients with major depressive disorder. Brain Behav Immun, 2015. 49: p. 206-15.
Lengvenyte, A., et al., Associations of white blood cell and platelet counts with specific depressive symptom dimensions in patients with bipolar disorder: Analysis of data from the FACE-BD cohort. Brain Behav Immun, 2023. 108: p. 176-187.
Chi, H., Immunometabolism at the intersection of metabolic signaling, cell fate, and systems immunology. Cell Mol Immunol, 2022. 19(3): p. 299-302.
Van Dyken, P. and B. Lacoste, Impact of Metabolic Syndrome on Neuroinflammation and the Blood-Brain Barrier. Front Neurosci, 2018. 12: p. 930.
Keyes, K.M. and D. Westreich, UK Biobank, big data, and the consequences of non-representativeness. Lancet, 2019. 393(10178): p. 1297.
Patient Health Questionnaire-4, Kroenke, K.. Spitzer, R. L., Williams, J. B. W., & Löwe, B.(2009). Patient Health Questionnaire-4 (PHQ-4) [Database record]. APA PsycTests.https://doi.org/10.1037/t06168-000.
The Corsini Encyclopedia of Psychology Full Access. Eysenck Personality Questionnaire. Nilly Mor. First published: 30 January 2010. https://doi.org/10.1002/9780470479216.corpsy0340Citations: 1. Fudan SFX
Jenkinson, M., et al., Fsl. Neuroimage, 2012. 62(2): p. 782-90.
Zhang, Z., Propensity score method: a non-parametric technique to reduce model dependence. Ann Transl Med, 2017. 5(1): p. 7.
Ormel, J., et al., Neuroticism and common mental disorders: meaning and utility of a complex relationship. Clin Psychol Rev, 2013. 33(5): p. 686-697.
Livingston, L.A., P. Shah, and F. Happe, Compensatory strategies below the behavioural surface in autism: a qualitative study. Lancet Psychiatry, 2019. 6(9): p. 766-777.
Allott, K., et al., Compensatory Interventions for Cognitive Impairments in Psychosis: A Systematic Review and Meta-Analysis. Schizophr Bull, 2020. 46(4): p. 869-883.
Livingston, L.A. and F. Happe, Conceptualising compensation in neurodevelopmental disorders: Reflections from autism spectrum disorder. Neurosci Biobehav Rev, 2017. 80: p. 729-742.
Kalayasiri, R., et al., Methamphetamine (MA) use and MA-induced psychosis are associated with increasing aberrations in the compensatory immunoregulatory system, interleukin-1alpha, and CCL5 levels. Transl Psychiatry, 2023. 13(1): p. 361.
Chen MH, et al., Association between psychiatric disorders and iron deficiency anemia among children and adolescents a nationwide population-based study. BMC Psychiatry, 2013. 13: p. 161.
Lee, H.S., et al., Psychiatric disorders risk in patients with iron deficiency anemia and association with iron supplementation medications: a nationwide database analysis. BMC Psychiatry, 2020. 20(1): p. 216.
Wang, Y., et al., Association between anemia and depression: results from NHANES 2005-2018 and mendelian randomization analyses. Ann Hematol, 2023. 102(10): p. 2651-2658.
Trevisan C., et al., Low Hemoglobin Levels and Risk of Developing Depression in the Elderly- Results From the Prospective PRO.V.A. Study. J Clin Psychiatry. 2016.77(12)-e1549-e1556.
Cheng, W., et al., Functional connectivity of the human amygdala in health and in depression. Soc Cogn Affect Neurosci, 2018. 13(6): p. 557-568.
Rolls, E.T., et al., Human amygdala compared to orbitofrontal cortex connectivity, and emotion. Prog Neurobiol, 2023. 220: p. 102385.
Rolls, E.T., W. Cheng, and J. Feng, The orbitofrontal cortex: reward, emotion and depression. Brain Commun, 2020. 2(2): p. fcaa196.
Cheng, W., et al., Medial reward and lateral non-reward orbitofrontal cortex circuits change in opposite directions in depression. Brain, 2016. 139(Pt 12): p. 3296-3309.
Rao, V.R., et al., Direct Electrical Stimulation of Lateral Orbitofrontal Cortex Acutely Improves Mood in Individuals with Symptoms of Depression. Curr Biol, 2018. 28(24): p. 3893-3902 e4.
Chen D, et al. Neurophysiological stratification of major depressive disorder by distinct trajectories. Nat Ment Heal. Springer US; 2023;1-863–875.
Jiang Y, Wang J, et al. Neuroimaging biomarkers define neurophysiological subtypes with distinct trajectories in schizophrenia. Nat Ment Heal. 2023;1-186–199.
Drevets, W.C., et al., Immune targets for therapeutic development in depression: towards precision medicine. Nat Rev Drug Discov, 2022. 21(3): p. 224-244.
Guzek, D., et al., Association between Vitamin D Supplementation and Mental Health in Healthy Adults: A Systematic Review. J Clin Med, 2021. 10(21).
Ronaldson, A., et al., Prospective associations between vitamin D and depression in middle-aged adults: findings from the UK Biobank cohort. Psychol Med, 2022. 52(10): p. 1866-1874.
Neriman, A., Y. Hakan, and U. Ozge, The psychotropic effect of vitamin D supplementation on schizophrenia symptoms. BMC Psychiatry, 2021. 21(1): p. 309.

There is NO Competing Interest.

SupplementaryAppendix.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

Temporal trends of blood-based markers in various mental disorders and their relationship with brain structure

Status:

Version 1

Abstract

Background:

Methods:

Results:

Conclusions:

Figures

Introduction

Methods

Study cohort

Outcome ascertainment

Blood-based biomarker measurements

Psychiatric assessments

Propensity score matching

Statistical analysis

Role of the funding source

Results

Study populations

Temporal trends of psychiatric assessments in mental disorders before and after clinical diagnosis

Temporal trends of blood-based makers in mental disorders before and after clinical diagnosis

Identified blood-based makers were associated with brain structures

Discussion

Strengths and Limitations

Conclusions

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1