Post-traumatic stress disorder (PTSD) is a syndrome resulting from exposure to actual or threatened serious injury, death or sexual assault (1). Post-traumatic stress disorder affects both civilians and active duty military personnel, although the latter are at increased risk, especially after combat tours. Pre-existing factors include gender, prior traumatic exposure, pre-existing mental illness, lower socio-economic status, lower intelligence and childhood adversity (2). Post-traumatic factors include development of acute stress disorder (ASD), other stresses such as financial problems, subsequent adverse life events and lack of social support. Surprisingly, exposure to traumatic events is common. In a recent study of 2,647 US adults exposed to a potentially traumatic event, 10.5% met the full criteria for lifetime PTSD however, 54.2% of the remaining non-PTSD group experienced trauma-related disturbances to sleep (3).

Psychiatric conditions such as PTSD are poorly understood and there is a wide heterogeneity in how the illness manifests in individuals. After a traumatic event, it is normal to have strong feelings of anxiety, sadness, or stress. Some individuals may experience nightmares, intrusive memories about the event, or problems sleeping, which are all common characteristics of PTSD. However, these symptoms do not necessarily mean that an individual has PTSD.

In 2013, the latest version of the Diagnostic and Statistical Manual of Mental Disorders (DSM) was released (DSM-5). The DSM is used by clinicians and psychiatrists to diagnose psychiatric illness. The DSM is focused on describing symptoms as well as statistics concerning which gender is most affected by the illness, age of onset and effect of treatment. The main issue with the DSM is validity (4). A statement issued by the National Institute for Mental Health (NIMH) stated that ‘the DSM-5 represented the best information currently available for clinical diagnosis of mental disorders.’ However, a closer examination of the DSM-5, reveals that its diagnoses are not accurate representations of mental disorders and they are not necessarily effective in determining what treatment approaches are best for the patient. A diagnosis is only valid when it accurately describes a patient’s condition or disorder. However, the diagnoses described in the DSM-5 are not objective medical conditions like how other conditions are described e.g. heart disease, diabetes, and cancer. Instead, they are symptoms and behaviours reported by patients which are interpreted by the clinician (5). Possible risks include misdiagnosis or even over diagnosis. Unfortunately, the clinician’s tendency to look for, find, and interpret these results, can result in confirmation bias. As such, patients can potentially be labelled as having a disorder simply because their behaviour does not always conform to the current ‘ideal’. In addition, patients are also at risk of stigmatisation (6).

To date, no clinically validated biomarker or biomarker combination has been found to aid in the diagnosis, treatment and management of patients with PTSD. However, the search for reliable and, possibly, specific biomarkers for mental disorders is an active area of research. Symptoms and comorbidity with other neuropsychiatric disorders e.g. depression and anxiety, suggest that a single biomarker is unlikely to be diagnostic. However, a combination of biomarkers, or biomarker signature, combined with the DSM-5, would allow clinicians to better manage patients who present with PTSD-like symptoms. The current invention provides combinations of biomarkers which, along with clinical risk factors, could be used to aid clinicians in identifying and/or stratifying patients suffering from or ‘at risk’ of PTSD or a comorbidity thereof. Surprisingly, it was also found that some biomarkers and combinations thereof could be used to differentiate patients with any mental health disorder from controls.

References

1. American Psychiatric Association. Diagnostic and Statistical Manual of Mental

Disorders [Internet] American Psychiatric Association; 2013 [cited 2019 Jul 4] Available from: https://psychiatryonline.Org/doi/book/10.1176/appi. books.9780890425596

2. Brewin CR, Andrews B, Valentine JD. Meta-analysis of risk factors for posttraumatic stress disorder in trauma-exposed adults. J Consult Clin Psychol [Internet] 2000 Oct [cited 2019 Jul 4];68(5):748-66. Available from: http://www.ncbi.nlm.nih.gOv/pubmed/11068961

3. Milanak ME, Zuromski KL, Cero I, Wilkerson AK, Resnick HS, Kilpatrick DG. Traumatic Event Exposure, Posttraumatic Stress Disorder, and Sleep Disturbances in a National Sample of U.S. Adults. J Trauma Stress [Internet] 2019 Feb [cited 2019 Jul 4];32(1): 14—22. Available from: http://doi.wiley.com/10.1002/jts.22360 4. Aboraya A, France C, Young J, Curci K, Lepage J. The Validity of Psychiatric Diagnosis Revisited: The Clinician’s Guide to Improve the Validity of Psychiatric Diagnosis. Psychiatry (Edgmont) [Internet] 2005 Sep [cited 2019 Jul 4];2(9):48— 55. Available from: http://www.ncbi.nlm.nih.gov/pubmed/21120108

5. Sisti D, Young M, Caplan A. Defining mental illnesses: can values and objectivity get along? BMC Psychiatry [Internet] 2013 Dec 24 [cited 2019 Jul 4];13(1):346. Available from : http://bmcpsychiatry. biomedcentral. com/articles/10.1186/1471 -244X-13-346

6. Mittal D, Drummond KL, Blevins D, Curran G, Corrigan P, Sullivan G. Stigma associated with PTSD: Perceptions of treatment seeking combat veterans. Psychiatr Rehabil J [Internet] 2013 [cited 2019 Jul 4]; 36 (2): 86-92. Available from:

7. R Core Team. R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing; 2018.

8. Moylan S, Jacka FN, Pasco JA, Berk M. How cigarette smoking may increase the risk of anxiety symptoms and anxiety disorders: a critical review of biological pathways. Brain Behav [Internet] 2013 May [cited 2019 Jul 4]; 3(3): 302-26. Available from: http://doi.wiley.eom/10.1002/brb3.137

9. Paulus EJ, Argo TR, Egge JA. The impact of posttraumatic stress disorder on blood pressure and heart rate in a veteran population. J Trauma Stress [Internet] 2013 Feb [cited 2019 Jul 4];26(1): 169-72. Available from: http://doi.wiley.com/10.1002/jts.21785

10. Hoffer M, Balaban C. Neurosensory Disorders in Mild Traumatic Brain Injury [Internet] 2018 [cited 2019 Jul 4] 247-277 p. Available from: https://books. google. co. uk/books?hl=en&lr=&id=fslaDwAAQBAJ&oi=fnd&pg =PP1&dq =neurosensory+disorders+in+mild+traumatic+brain+injury+chapt er+16+citation&ots=ii ROw6BPLV&sig=OnyAdMmz8FJ5WqVIDOvxTkdOx7U

11. Yang Z, Wang KKW. Glial fibrillary acidic protein: from intermediate filament assembly and gliosis to neurobiomarker. Trends Neurosci [Internet] 2015 Jun [cited 2019 Jul 4];38(6):364-74. Available from: https://linkinghub.elsevier.com/retrieve/pii/S01662236150008 18

12. Okonkwo DO, Yue JK, Puccio AM, Panczykowski DM, Inoue T, McMahon PJ, et al. GFAP-BDP as an acute diagnostic marker in traumatic brain injury: results from the prospective transforming research and clinical knowledge in traumatic brain injury study. J Neurotrauma [Internet] 2013 Sep 1 [cited 2019 Jul 4];30(17):1490-7. Available from: http://www.liebertpub.eom/doi/10.1089/neu.2013.2883

Summary of the invention

In a first embodiment, the current invention provides a method for aiding in the diagnosis of a mental disorder in a patient suspected of suffering therefrom, said method comprising; a) determining the level of a biomarker selected from the list consisting of Glial fibrillary acidic protein (GFAP), neuron-specific enolase (NSE), midkine, folate, homocysteine, high-density lipoprotein (HDL) cholesterol, ferritin, adiponectin, interleukin 8 (IL-8), monocyte chemoattractant protein 1 (MCP-1), plasminogen activator inhibitor-1 (PAI-1), Vascular endothelial growth factor (VEGF), epidermal growth factor (EGF), tumour necrosis factor alpha (TNFa), Cystatin C, the fibrin degradation product D-dimer, neutrophil gelatinase-associated lipocalin (NGAL) and Brain-derived neurotrophic factor (BDNF)in an ex vivo sample isolated from the patient, b) establishing the significance of the level of the biomarker.

In a second embodiment, the current invention provides a method for aiding in the diagnosis of a mental disorder in a patient suspected of suffering therefrom, said method comprising; a) determining the level of two or more biomarkers selected from the list consisting of GFAP, NSE, midkine, folate, vitamin BI ₂ , iron, homocysteine, HDL cholesterol, LDL cholesterol, total cholesterol, C-reactive protein (CRP), ferritin, insulin, leptin, adiponectin, cystatin C, PAI-1 , tissue plasminogen activator (tPA), PAI-1 /tPA complex, resistin, interleukin 2 (IL-2), interleukin 4 (IL-4), interleukin 6 (IL- 6), IL-8, interleukin 10 (IL-10), VEGF, interferon gamma (I FN y), TNFa, interleukin 1 alpha (I L-1 a), interleukin 1 beta (I L-1 b), MCP-1 , EGF, D-dimer, NGAL, Soluble tumour necrosis factor receptor-1 (sTNFRI), BDNF and heart-type fatty acid binding protein (H-FABP)in an ex vivo sample isolated from the patient, b) establishing the significance of the level of biomarkers. Preferred mental disorders include PTSD, depression, anxiety, panic disorder and substance abuse. Preferred biomarker combinations include; i) GFAP + NSE ii) GFAP + folate iii) GFAP + HDL cholesterol iv) GFAP + HDL cholesterol + LDL cholesterol v) GFAP + NSE + d-dimer vi) GFAP + NSE + Cystatin C vii) GFAP + NSE + Folate viii)GFAP + NSE + HDL cholesterol ix) GFAP + NSE + BDNF + Cystatin C + folate x) GFAP + BDNF + D-dimer + Cystatin C + HDL cholesterol + LDL cholesterol + folate

When the mental disorder is PTSD preferred biomarker combinations include; i) NSE + GFAP ii) Ferritin + NSE + GFAP iii) Ferritin + NSE + VEGF + GFAP iv) HDL cholesterol + VEGF + ferritin v) HDL cholesterol + NSE + VEGF + GFAP vi) Ferritin, HDL cholesterol, NSE vii) Ferritin, HDL cholesterol, GFAP

In a further embodiment, the invention provides methods for aiding the diagnosis of post-traumatic stress disorder (PTSD) in a male patient suspected of suffering therefrom, wherein one of the following combinations of biomarkers is selected - i) HDL cholesterol + ferritin + NSE + IL-2 ii) Ferritin + IL-2 + NSE iii) Ferritin + NSE iv) HDL cholesterol, ferritin, cystatin C, IL-2, IL-8, VEGF, D-dimer and NSE

Further still, the invention provides methods for aiding the diagnosis of post- traumatic stress disorder (PTSD) in a female patient suspected of suffering therefrom, wherein one of the following combinations of biomarkers is selected - i) HDL cholesterol + GFAP + MCP-1 ii) GFAP + MCP-1 iii) HDL cholesterol, ferritin, insulin, adiponectin, IL-8, VEGF, TNFa, MCP-1 , EGF, D- dimer, NSE, NGAL, BDNF and GFAP The current invention also provides methods for aiding in the diagnosis of a comorbidity in a PTSD patient, said methods comprising; a) determining the level of two or more biomarkers selected from the list consisting of GFAP, NSE, midkine, folate, vitamin B12, iron, homocysteine, HDL cholesterol, ferritin, LDL cholesterol, adiponectin, IL1a, IL-2, IL-4, IL-8, MCP-1 , PAMtPA, VEGF, EGF, Cystatin C, D- dimer, NGAL and BDNF in an ex vivo sample isolated from a patient, b) establishing the statistical significance of the level of the biomarkers. These methods include; i) A method for aiding in the diagnosis of depression in a PTSD patient, said method comprising a) determining the level of GFAP and HDL cholesterol in an ex vivo sample taken from the patient, b) establishing the significance of the level of biomarkers. ii) A method for aiding in the diagnosis of panic disorder in a PTSD patient, said method comprising a) determining the level of GFAP and IL1 a in an ex vivo sample taken from the patient, b) establishing the significance of the level of biomarkers. iii) A method for aiding in the diagnosis of anxiety in a PTSD patient, said method comprising a) determining the level of GFAP, HDL cholesterol and leptin in an ex vivo sample taken from the patient, b) establishing the significance of the level of biomarkers. iv) A method for detecting substance abuse in a PTSD patient, said method comprising a) determining the level of HDL cholesterol, PAMtPA and IL-4 in an ex vivo sample taken from the patient, b) establishing the significance of the level of biomarkers.

Also provided are solid-state devices which enable the level of biomarkers to be determined.

Description of the figures

Figure 1 - Serum neuron-specific enolase (NSE) (ng/ml) levels for control vs. PTSD participants ( ^**** : p <= 0.0001) (all genders)

Figure 2 - Serum high density lipoprotein (HDL) cholesterol (mmol/l) levels for control vs. PTSD participants ( ^**** : p <= 0.0001) (all genders) Figure 3 - Serum vascular endothelial growth factor (VEGF) (pg/ml) levels for control vs. PTSD participants ( ^**** : p <= 0.0001) (all genders)

Figure 4 - Serum glial fibrillary acidic protein (GFAP) (ng/ml) levels for control vs. PTSD participants ( ^**** : p <= 0.0001) (all genders) Figure 5 - Serum ferritin (mg/I) levels for control vs. PTSD participants ( ^*** : p <= 0.001) (all genders)

Figure 6 - Serum folate (ug/L) levels for control vs. PTSD participants ( ^**** : p <= 0.0001) (all genders)

Figure 7 - Serum homocysteine (pmol/L) levels for control vs. PTSD participants ( ^* : p <=0.05) (all genders)

Figure 8 - Serum ferritin levels (mg/I) for control female, PTSD female, control male and PTSD male study participants ( ^* : p <= 0.05, ^** : p <= 0.01 , ^**** : p <= 0.0001) (all genders)

Figure 9 - Serum IL-2 (pg/ml) for male controls vs male PTSD participants ( ^* : p <= 0.05)

Figure 10 - Serum MCP-1 (pg/ml) for female controls vs. female PTSD participants ( _**** . _{p <=} 0.0001)

Figure 11 - Area under the receiver operator curve for individual biomarker and biomarker combination algorithms used to differentiate between control vs. PTSD participants (all genders)

Figure 12 - Area under the receiver operator curve for individual biomarker and biomarker combination algorithms used to differentiate between control vs. PTSD participants (all genders)

Figure 13 - Area under the receiver operator curve for individual biomarker and biomarker combination algorithms used to differentiate between control vs. PTSD participants (all genders)

Figure 14 - Area under the receiver operator curve for individual biomarker and biomarker combination algorithms used to differentiate between control vs. PTSD participants (males) Figure 15 - Area under the receiver operator curve for individual biomarker and biomarker combination algorithms used to differentiate between control vs. PTSD participants (males)

Figure 16 - Area under the receiver operator curve for individual biomarker and biomarker combination algorithms used to differentiate between control vs. PTSD participants (females)

Figure 17 - Area under the receiver operator curve for individual biomarker and biomarker combination algorithms used to differentiate between control vs. PTSD participants (females) Figure 18 - Area under the receiver operator curve for the combination of GFAP and NSE used to differentiate between controls and PTSD (AUC 0.865) and also between controls and all mental disorders (AUC 0.908) (all genders)

Detailed description of the invention The present invention describes biomarker-based methods to aid in the diagnosis of a mental disorder. Specifically, the methods comprise measurement of relative levels or concentrations of biomarkers in ex vivo samples taken from a patient. It is further envisaged that the invention may also be used for monitoring the progression or recurrence of a mental disorder and the effectiveness of any treatment strategy which has been implemented. Although not proven herein, there is also potential that the biomarker combinations disclosed could help to determine persons at risk of developing a mental disorder.

The term “patient” refers to any person presenting with symptoms suggestive of a mental disorder. The patient may be a person presenting for routine screening or they may present with symptoms suggestive of a mental disorder. The patient may also be an individual deemed at high risk for a mental disorder, due to having experienced a traumatic event or family history for example. Alternatively, the patient could be an individual who has received treatment for a mental disorder and the screen is to monitor progress or detect possible recurrence. The term “mental disorder” as used herein refers to any behavioural or mental pattern which causes distress or impairment of personal functioning. These include any anxiety disorders, eating disorders, mood disorders, neurodevelopmental disorders, personality disorders, psychotic disorders and substance use disorders. Preferred mental disorders include PTSD, depression, anxiety, panic disorder and substance abuse. Even more preferably PTSD and comorbidities thereof.

The term “biomarker”, in the context of the current invention, refers to a molecule present in a biological sample of a patient, the levels of which may be indicative of a mental disorder. Such molecules may include peptides/proteins or nucleic acids and derivatives thereof. As used herein the term ‘determining’ means quantitatively analysing for the amount of a substance present, in this case the biomarkers in a patient sample.

In a first embodiment, the current invention provides a method for aiding in the diagnosis of a mental disorder in a patient suspected of suffering therefrom, said method comprising; a) determining the level of a biomarker selected from the list consisting of Glial fibrillary acidic protein (GFAP), neuron-specific enolase (NSE), midkine, folate, homocysteine, high-density lipoprotein (HDL) cholesterol, ferritin, adiponectin, interleukin 8 (IL-8), monocyte chemoattractant protein 1 (MCP-1), plasminogen activator inhibitor-1 (PAI-1), Vascular endothelial growth factor (VEGF), epidermal growth factor (EGF), tumour necrosis factor alpha (TNFa), Cystatin C, the fibrin degradation product D-dimer, neutrophil gelatinase-associated lipocalin (NGAL) and Brain-derived neurotrophic factor (BDNF) in an ex vivo sample isolated from the patient, b) establishing the significance of the level of the biomarker. Preferably the biomarker is selected from GFAP, NSE, HDL cholesterol, VEGF, folate, d-dimer, MCP-1 and ferritin. Even more preferably GFAP or NSE.

In a second aspect of the present invention, combinations of two or more biomarkers can be used in the method to aid in the diagnosis of a mental disorder, the two or more biomarkers being selected from the list consisting of GFAP, NSE, midkine, folate, vitamin BI ₂ , iron, homocysteine, HDL cholesterol, LDL cholesterol, total cholesterol, C-reactive protein (CRP), ferritin, insulin, leptin, adiponectin, cystatin C, PAI-1 , tissue plasminogen activator (tPA), PAI-1/tPA complex, resistin, interleukin 2 (IL-2), interleukin 4 (IL-4), interleukin 6 (IL-6), IL-8, interleukin 10 (IL-10), VEGF, interferon gamma (IFNy), TNFa, interleukin 1 alpha (IL-1 a), interleukin 1 beta (IL-1 b), MCP-1 , EGF, D-dimer, NGAL, Soluble tumour necrosis factor receptor-1 (sTNFRI), BDNF and heart-type fatty acid binding protein (H-FABP).Most preferably the two or more biomarkers are selected from the list consisting of GFAP, NSE, midkine, folate, homocysteine HDL cholesterol, ferritin, LDL cholesterol, adiponectin, IL-8, MCP-1 , EGF, D-dimer, NGAL and BDNF. Preferred mental disorders include PTSD, depression, anxiety, panic disorder and substance abuse. Preferred biomarker combinations include; i) GFAP + NSE ii) GFAP + folate iii) GFAP + HDL cholesterol iv) GFAP + HDL cholesterol + LDL cholesterol v) GFAP + NSE + d-dimer vi) GFAP + NSE + Cystatin C vii) GFAP + NSE + Folate viii) GFAP + NSE + HDL cholesterol ix) GFAP + NSE + BDNF + Cystatin C + folate x) GFAP + BDNF + D-dimer + Cystatin C + HDL cholesterol + LDL cholesterol + folate

As well as determining if a patient is suffering from a mental disorder or not, the current invention also provides biomarker combinations which aid in the diagnosis of specific mental disorders. For example, when the mental disorder is PTSD preferred biomarker combinations include; i) NSE + GFAP ii) Ferritin + NSE + GFAP iii) Ferritin + NSE + VEGF + GFAP iv) HDL cholesterol + VEGF + ferritin v) HDL cholesterol + NSE + VEGF + GFAP vi) Ferritin, HDL cholesterol, NSE vii) Ferritin, HDL cholesterol, GFAP viii) HDL cholesterol, LDL cholesterol, ferritin, adiponectin, IL-8, VEGF, MCP-1 , EGF, D-dimer, NSE, NGAL, BDNF, GFAP

An analysis of the biomarker data for a subset of PTSD cases versus a subset of controls from whom a PCL5 score was available resulted in combinations with high discriminatory ability (table 15). Therefore, most preferably the invention provides a method to aid in the diagnosis of post-traumatic stress disorder (PTSD) in a patient suspected of suffering therefrom, said method comprising a) determining the level of two or more biomarkers selected from plasminogen activator inhibitor-1 (PAI-1), interleukin 8 (IL-8), cystatin C, tissue plasminogen activator (tPA), PAI-1 /tPA complex, interleukin 4 (IL-4), interferon gamma (IFNy) and epidermal growth factor (EGF), b) establishing the significance of the level of the biomarkers by inputting each of the biomarker concentration values into a statistical methodology to produce an output value that indicates whether the patient has or is at risk of developing PTSD. In preferred embodiments at least one biomarker is selected from the group consisting of PAI-1 , tPA and PAI-1/tPA complex and at least one biomarker is selected from the group consisting of IL-8, cystatin C, IL-4, IFNy and EGF. Combinations with the best discriminatory ability included: i) PAI-1 + IL-8 ii) tPA + Cystatin C iii) tPA + PAI-1 iv) EGF+ tPA v) PAI-1 + Cystatin C vi) IL-4 + IL-8 vii) PAI-1 + IL-4 viii) tPA + PAI-1 + Cystatin C ix) PAI-1 + IL-4 + IL-8 xi) IFNy + IL-8 + PAI-1 -tPA + tPA + PAI-1 + Cystatin C xii) EGF+ IFNY+ IL-8 + PAI-1-tPA + tPA + PAI-1 + Cystatin C. In addition to any of the combinations listed above it is within the scope of the invention to determine levels of further biomarkers which could contribute to a diagnosis of PTSD, for example, but not limited to, midkine, folate, vitamin BI ₂ , iron, homocysteine, HDL cholesterol, LDL cholesterol, total cholesterol, CRP, ferritin, insulin, leptin, adiponectin, cystatin C, PAI-1 , tPA, PAI-1/tPA, resistin, IL-2, IL-4, IL-6, IL-8, IL-10, VEGF, IFNy, TNFa, IL-1a, IL-1 b, MCP-1 , EGF, D-dimer, NSE, NGAL, STNFR1 , BDNF and H-FABP.

If the subjects of the invention are divided by gender then additional biomarker combinations to aid in the diagnosis of PTSD are preferred, for example if the subject is male two or more biomarkers are selected from ferritin, NSE, GFAP, midkine, HDL cholesterol, cystatin C, IL-2, IL-8, VEGF and D-dimer. In preferred embodiments for male patients one of the following combinations of biomarkers is selected to aid in the diagnosis of PTSD- i) Ferritin + NSE ii) GFAP + NSE iii) Ferritin + IL-2 + NSE iv) HDL cholesterol + ferritin + NSE + IL-2 v) HDL cholesterol, ferritin, cystatin C, IL-2, IL-8, VEGF, D-dimer and NSE

In another example, if the subject is female two or more biomarkers are selected from GFAP, MCP-1 , NSE, folate, HDL cholesterol, ferritin, insulin, adiponectin, IL-8, VEGF, TNFa, EGF, D-dimer, NGAL and BDNF. In preferred embodiments for female patients one of the following combinations of biomarkers is selected to aid in the diagnosis of PTSD - i) GFAP + MCP-1 ii) GFAP + NSE iii) HDL cholesterol + GFAP + MCP-1 iv) HDL cholesterol, ferritin, insulin, adiponectin, IL-8, VEGF, TNFa, MCP-1 , EGF, D-dimer, NSE, NGAL, BDNF and GFAP In the context of the present invention, a deviation from a control value for each of the biomarkers may be an indication that the patient suffers from a mental disorder. Dependent on the individual biomarker this deviation may be an increase or a decrease from a control value. For example, in the patient cohort of the current invention, levels of HDL cholesterol, VEGF and NSE were higher in patients with a mental disorder when compared to those in controls. Levels of D-dimer were lower in patients with a mental disorder than in controls (table 5). In the analysis which was restricted to controls for whom a PCL5 score had been assigned, EGF, PAI-1 , D- dimer, cystatin C and homocysteine were significantly lower in the PTSD group. tPA, IL-8, midkine, folic acid and INFy were significantly elevated in PTSD participants. The results for cystatin C are of particular interest as in the prior art higher levels than controls have been found in subjects with depression. Opposite expression profiles for tPA and PAI-1 in PTSD participants in the current invention compared to prior art for subjects with depression were also obsessed. These biomarkers may therefore have utility in differentiating PTSD from depression.

The current invention provides biomarker combinations which aid in the diagnosis of a mental disorder. Early detection and treatment of mental disorders is critical and can dramatically increase the patients’ quality of life. Additionally, the biomarker combinations of the current invention allow the monitoring of a mental disorder within an individual through serial testing of samples from said individual over an extended period. For example, routine determination of the levels of the biomarkers of the preferred combinations could detect the changes from control values, which are indicative of the development of a mental disorder. A further use of the biomarker combinations of the current invention may be in predicting the risk of developing a mental disorder, particularly PTSD, in individuals who have suffered a traumatic event, although confirmation of this utility will require further studies.

A further aspect of the present invention is a method of determining the efficacy of a treatment for a mental disorder comprising determining the levels of two or more biomarkers selected from the list consisting of midkine, folate, vitamin BI ₂ , iron, homocysteine, HDL cholesterol, LDL cholesterol, total cholesterol, CRP, ferritin, insulin, leptin, adiponectin, cystatin C, PAI-1 , tPA, PAI-1/tPA, resistin, IL-2, IL-4, IL-6, IL-8, IL-10, VEGF, IFNy, TNFa, IL-1a, IL-1 b, MCP-1 , EGF, D-dimer, NSE, NGAL, sTNFRI , BDNF, HFABP and GFAP in a sample from a treated patient, and comparing levels with those from a healthy control or with levels from the same patient before treatment, wherein dependent on the biomarker, either an increase or decrease in level indicates the effectiveness of the treatment. Most preferably the two or more biomarkers are selected from the list consisting of GFAP, NSE, HDL cholesterol, ferritin, folate, homocysteine, LDL cholesterol, adiponectin, IL-8, MCP-1 , EGF, D-dimer, NGAL and BDNF. The treatment can be for example, a drug treatment or appropriate psychiatric care such as cognitive behavioural therapy. In a preferred method for determining the efficacy of a treatment for a mental disorder, one of the two or more markers is GFAP or NSE. Wherein the treatment is a drug treatment, the method of determining the efficacy of the drug treatment for the mental disorder would comprise determining the levels of biomarkers, for example GFAP and NSE in a sample from a patient treated with the drug, and comparing biomarker levels with those from a healthy control or with levels from the same patient before treatment with the drug, wherein, dependent on the biomarker, either an increase or decrease in level indicates the effectiveness of the drug treatment.

In addition to methods to aid in the diagnosis of PTSD, the current invention also provides methods for aiding in the diagnosis of comorbidities thereof. These include; i) A method for aiding in the diagnosis of depression in a PTSD patient, said method comprising a) determining the level of two or more biomarkers selected from GFAP, HDL cholesterol, adiponectin, VEGF, TNFa, MCP-1 , EGF, D-dimer, NSE, NGAL and BDNF and in an ex vivo sample taken from the patient, b) establishing the significance of the level of biomarkers. Preferably the levels of HDL cholesterol and GFAP are determined. ii) A method for aiding in the diagnosis of panic disorder in a PTSD patient, said method comprising a) determining the level of two or more biomarkers selected from GFAP, I L-1 a, CRP, insulin, INFy and BDNF in an ex vivo sample taken from the patient, b) establishing the significance of the level of biomarkers. Preferably the levels of GFAP and I L-1 a are determined. iii) A method for aiding in the diagnosis of anxiety in a PTSD patient, said method comprising a) determining the level of two or more biomarkers selected from GFAP, HDL cholesterol, leptin, insulin, LDL cholesterol, cystatin C, IL-6, D-dimer, BDNF and ferritin in an ex vivo sample taken from the patient, b) establishing the significance of the level of biomarkers. Preferably the levels of HDL cholesterol, GFAP and leptin are determined. iv) A method for detecting substance abuse in a PTSD patient, said method comprising a) determining the level of two or more biomarkers selected from HDL cholesterol, PAH/tPA, IL-4, IL-8, IL-10, tPA, total cholesterol and TNFa in an ex vivo sample taken from the patient, b) establishing the significance of the level of biomarkers. Preferably the levels of HDL cholesterol, PAI-1/tPA and IL-4 are determined.

The methods for detecting comorbidities in PTSD patients described above may also have utility in detecting these comorbidities in non-PTSD populations. For example, the depression and anxiety groups included in the analysis were composed of PTSD patients with these conditions as well as a small number of controls with each. There was a trend towards higher GFAP in these control individuals with depression and anxiety, future studies will determine whether this is significant and GFAP is useful in aiding the diagnosis of these conditions in non-PTSD populations.

The methods of the current invention are intended to compliment, and indeed enhance, current diagnostic tools such as the DSM-5 criteria for mental disorders and comorbidities thereof. While biomarker combinations provided herein may be used in isolation, it is envisaged that they will be of most benefit when used along with clinical data and/or a suitable classification system such as the DSM-5 or an equivalent. For example, the biomarkers can indicate that the patient is suffering from a mental disorder and then clinical data or the classification system can determine which mental disorder they have. Alternatively, the biomarkers can be used to confirm the diagnosis of a particular mental disorder, most preferably PTSD.

In the context of the present invention, a “control value” is understood to be the level of a biomarker, such as GFAP, NSE, ferritin, VEGF or HDL cholesterol typically found in individuals not suffering from a mental disorder. The control level of a biomarker may be determined by analysis of a sample isolated from an individual (i.e. individuals not suffering from or developing a disease or disorder, preferably a mental disorder) or may be the level of the biomarker understood by the skilled person to be typical for a healthy individual. The control value may be a range of values considered by the skilled person to be a normal level for the biomarker in a healthy individual or person not suffering from a mental disorder. The skilled person will appreciate that control values for a biomarker may be calculated by the user analysing the level of the biomarker in a sample from a healthy individual, a person not suffering from a mental disorder or by reference to typical values provided by the manufacturer of the assay used to determine the level of a biomarker in the sample.

The “sample” of the current invention can be any ex vivo biological sample from which the levels of biomarkers can be determined. Preferably, the sample isolated from the patient is a serum or plasma sample. However, the sample could be selected from, for example, whole blood, urine, saliva or sputum. The determination of the level of biomarkers may be carried out on one or more samples taken from the patient. The sample may be obtained from the patient by methods routinely used in the art.

The determination of the level of biomarkers in the sample may be determined by immunological methods such as an ELISA-based assay. The methods of the current invention preferably comprise the following steps; the biomarkers binding to a probe(s), adding a detector probe(s) and detecting and measuring the biomarker/probe complex signal(s), placing these values into a machine algorithm and analysing the output value, said value indicating whether the patient has or is at risk of having a mental disorder. Preferably, the methods of the present invention use a solid-state device for determining the level of biomarkers in the sample isolated from the patient.

The solid-state device comprises a substrate having a probe or multiple different probes immobilised upon it that bind specifically to a biomarker. The interactions between a biomarker and its respective probe can be monitored and quantified using various techniques that are well-known in the art. The term “probe” refers to a molecule that is capable of specifically binding to a target molecule, in this case the biomarkers, such that the target molecule can be detected as a consequence of said specific binding. Probes that can be used in the present invention include, for example, antibodies, aptamers, phages and oligonucleotides. In a preferred embodiment of the current invention the probe is an antibody. The “level” of a biomarker refers to the amount, expression level or concentration of the biomarker within the sample. This level can be a relative level in comparison to another biomarker or a previous sample.

The term “antibody” refers to an immunoglobulin which specifically recognises an epitope on a target as determined by the binding characteristics of the immunoglobulin variable domains of the heavy and light chains (VHS and VLS), more specifically the complementarity-determining regions (CDRs). Many potential antibody forms are known in the art, which may include, but are not limited to, a plurality of intact monoclonal antibodies or polyclonal mixtures comprising intact monoclonal antibodies, antibody fragments (for example Fab, Fab’, and Fv fragments, linear antibodies single chain antibodies and multispecific antibodies comprising antibody fragments), single-chain variable fragments (scFvs), multi specific antibodies, chimeric antibodies, humanised antibodies and fusion proteins comprising the domains necessary for the recognition of a given epitope on a target. Preferably, references to antibodies in the context of the present invention refer to polyclonal or monoclonal antibodies. Antibodies may also be conjugated to various reporter moieties for a diagnostic effect, including but not limited to radionuclides, fluorophores, dyes or enzymes including, for example, horse-radish peroxidase and alkaline phosphatase.

Such antibodies may be immobilised at discrete areas of an activated surface of the substrate. The solid-state device may perform multi-analyte assays such that the level of a biomarker in a sample isolated from the patient may be determined simultaneously with the level of a further biomarker of interest in the sample. In this embodiment, the solid-state device has a multiplicity of discrete reaction sites each bearing a desired antibody covalently bound to the substrate, and in which the surface of the substrate between the reaction sites is inert with respect to the target biomarker. The solid-state, multi-analyte device may therefore exhibit little or no non specific binding. The combination of biomarkers may also be referred to as a panel of biomarkers.

The substrate can be any surface able to support one or more probes, but is preferably a biochip. A biochip is a planar substrate that may be, for example, mineral or polymer based, but is preferably ceramic. When identifying the various biomarkers/proteins of the invention it will be apparent to the skilled person that as well as identifying the full-length protein, the identification of a fragment or several fragments of a protein is possible, provided this allows accurate identification of the protein. Similarly, although a preferred probe of the invention is a polyclonal or monoclonal antibody, other probes such as aptamers, molecular imprinted polymers, phages, short chain antibody fragments and other antibody-based probes may be used.

A solid-state device that may be used in the invention may be prepared by activating the surface of a suitable substrate and applying an array of antibodies on to the discrete sites on the surface. If desired, the other active areas may be blocked. The ligands may be bound to the substrate via a linker. In particular, it is preferred that the activated surface is reacted successively with an organosilane, a bi-functional linker and the antibody. The solid-state device used in the methods of the present invention may be manufactured according to the method disclosed in, for example, GB-A-2324866 the contents of which are incorporated herein in its entirety. The solid-state device can be any substrate to which probes of the current invention can be attached for example a microtitre plate or beads. Preferably, the solid-state device used in the methods of the present invention is a biochip. The biochip may be a biochip which is incorporated into the Biochip Array Technology System (BAT) available from Randox Laboratories Limited (Crumlin, UK).

Preferably, a solid-state device may be used to determine the levels of two or more biomarkers selected from the list comprising GFAP, NSE, midkine, folate, vitamin Bi2, iron, homocysteine HDL cholesterol, LDL cholesterol, total cholesterol, CRP, ferritin, insulin, leptin, adiponectin, cystatin C, PAI-1, tPA, PAI-1/tPA, resistin, IL-2, IL-4, IL-6, IL-8, IL-10, VEGF, IFNy, TNFa, IL-1a, IL-1 b, MCP-1 , EGF, D-dimer, NGAL, sTNFRI, BDNF and H-FABP in the sample isolated from the patient. Most preferably the two or more biomarkers are selected from the list consisting of GFAP, NSE, folate, homocysteine, HDL cholesterol, ferritin, LDL cholesterol, adiponectin, IL-8, MCP-1 , EGF, D-dimer, NGAL and BDNF. The biomarkers to be determined using a solid-state device may be any of the combinations listed in the sections above. In a preferred embodiment of the current invention one of the two or more biomarkers is GFAP or NSE. In a preferred embodiment, the solid-state device comprises a substrate having an activated surface on to which is applied antibodies specific to each of the two or more biomarkers to discrete areas of the activated surface. Therefore, the solid-state device may perform multi-analyte assays such that the levels of biomarkers, for example GFAP, ferritin and NSE, in a sample may be determined simultaneously. In this embodiment, the solid-state device has a multiplicity of discrete reaction sites each bearing a desired antibody covalently bound to the substrate, and in which the surface of the substrate between the reaction sites is inert with respect to the target biomarkers. Each probe, whether individually or in multiplex, is specific to one target analyte. For example, a probe to NSE will only show specific binding to this analyte and will have no significant cross reactivity with HDL cholesterol, LDL cholesterol, total cholesterol, CRP, ferritin, insulin, leptin, adiponectin, cystatin C, PAI-1 , tPA, PAI-1/tPA, resistin, IL-2, IL-4, IL-6, IL-8, IL-10, VEGF, IFNy, TNFa, IL-1a, II_-1b, MCP-1 , EGF, D-dimer, NGAL, sTNFRI , BDNF, HFABP and GFAP, or indeed any other potentially interfering substance which could compromise the assay. The solid-state device not only has potential in diagnosis but also in monitoring the progression and risk of developing a mental disorder as well as determining the success of any treatments.

When the solid-state device is to be used to aid in the diagnosis of PTSD the preferred biomarkers are selected from PAI-1 , IL-8, cystatin C, tPA, PAI-1/tPA complex, IL-4, IFNy and EGF. More preferably at least one biomarker is selected from the group consisting of PAI-1 , tPA and PAI-1/tPA complex and at least one biomarker is selected from the group consisting of IL-8, cystatin C, IL-4, IFNy and EGF. Suitable combinations for a solid-state device to aid in the diagnosis of PTSD include - PAI-1 + IL-8, tPA + Cystatin C, tPA + PAI-1 , EGF+ tPA, PAI-1 + Cystatin C, IL-4 + IL-8, PAI-1 + IL-4, tPA + PAI-1 + Cystatin C, PAI-1 + IL-4 + IL-8, IFNy + IL-8 + PAI-1 -tPA + tPA + PAI-1 + Cystatin C and EGF+ IFNy+ IL-8 + PAI-1 -tPA + tPA + PAI-1 + Cystatin C.

In a preferred embodiment of the current invention each of the biomarker concentration values is inputted into a statistical methodology to produce an output value that correlates with the chances that the patient has or is at risk of developing a mental disorder. Preferably, the statistical methodology used is logistic regression, decision trees, support vector machines, neural networks, random forest or another machine-learning algorithm. The performance of the results of the applied statistical methods used in accordance with the present invention can be best described by their receiver operating characteristics (ROC). The ROC curve addresses both the sensitivity (the number of true positives) and the specificity (the number of true negatives) of the test. Therefore, sensitivity and specificity values for a given combination of biomarkers are an indication of the performance of the test. For example, if a biomarker combination has a sensitivity and specificity value of 80%, out of 100 patients, 80 will be correctly identified from the determination of the presence of the combination of biomarkers as positive for the disease, while out of 100 patients who do not have the disease 80 will accurately test negative for the disease.

A suitable statistical classification model, such as logistic regression, can be derived for a combination of biomarkers. Moreover, the logistic regression equation can be extended to include other (clinical) variables such as age and gender of the patient as well. In the same manner as described before, the ROC curve can be used to access the performance of the discrimination between patients and controls by the logistic regression model. Therefore, the logistic regression equation can be used apart or combined with other clinical characteristics to aid clinical decision making. Although a logistic regression equation is a common statistical procedure used in such cases and is preferred in the context of the current invention, other mathematical/statistical, decision trees or machine learning procedures can also be used.

One convenient goal to quantify the diagnostic accuracy of a laboratory test is to express its performance by a single number. The most common global measure is the area under the curve (AUC) of the ROC plot. The area under the ROC curve is a measure of the probability that the perceived measurement will allow correct identification of a condition. Values typically range between 1.0 (perfect separation of the test values of the two groups) and 0.5 (no apparent distributional difference between the two groups of test values). The area does not depend only on a particular portion of the plot such as the point closest to the diagonal or the sensitivity at 90% specificity, but on the entire plot. This is a quantitative, descriptive expression of how close the ROC plot is to the perfect one (area = 1.0). In the context of the present invention, the two different conditions can be whether a patient has or does not have a mental disorder. Methods and Materials

Patients

In total, n=78 participants recruited in the US by Discovery Life Sciences (1236 Los Osos Valley Rd, Suite T, Los Osos, CA93402 USA) and PrecisionMed (132 N. Arcia Ave, Solana Beach, CA92075, USA) were involved in this age and sex matched case-control study (mean age 39.5±10.2, median age 38.5; range 19 -70 years). A diagnosis of PTSD was positive if the mental health professional combined information about frequency and intensity of an individual’s experience into a single severity score. Clinician-administered PTSD scale for DSM-5 (CAPS-5) total symptom severity score is calculated by summing severity score for the 20 DSM-5 PTSD symptoms. A CAPS-5 or PTSD checklist for DSM-5 (PCL-5) score of <33 was considered negative for PTSD. The data in this study is presented using the assumption that both CAPS-5 and PCL-5 methodologies are the same. Venous blood samples were obtained from each participant along with a detailed clinical history. Thirty-nine patients were clinically diagnosed (CAPS-5) with PTSD (mean age 41.9±12.1 , median 41.0; range 23 - 70 years). Study participants included Caucasians (n=39 (50%)), Black African (n=25 (32%)), Hispanic or Latino (n=13 (16.7%)), and one (1.3%) participant with their ethnicity not recorded. The study conformed to all Data Use Agreements.

Clinical characteristics

Socio-demographic and clinical factors were collected from each participant to include; age, gender, ethnicity, tobacco and alcohol use, BMI, pulse, blood pressure, current medications, pain score (current, average, chronic), sleep disturbances and comorbidities e.g. depression, anxiety, panic disorder, Attention Deficit Disorder (ADD), bipolar disease, substance abuse, migraine, diabetes, and hypertension (Table 1).

Sampling and laboratory methods Analyses of biomarkers was completed using cytokine arrays (Randox Laboratories Ltd, Crumlin, UK) for the following proteins: Cytokine I Array: lnterleukin-1 a, -1 b, -2, - 4, -6, -8, -10, VEGF, EGF, TNFa, IFNy, and MCP-1; Metabolic Array I: Ferritin, insulin, leptin, plasminogen activator inhibitor-1 (PAI-1), and resistin; Metabolic Array II: C- reactive protein (CRP), adiponectin and cystatin C; Cerebral Array I: Brain-derived neurotropic factor (BDNF), glial fibrillary acidic protein (GFAP), and heart-type fatty acid binding protein (H-FABP); Cerebral Array II: D-dimer, neuron specific enolase (NSE), neutrophil gelatinase-associated lipocalin (NGAL), and soluble tumour necrosis factor receptor I (sTNFRI). Arrays were run on Evidence Investigator ^® analysers according to manufacturer’s instructions (Randox Laboratories Ltd, Crumlin, UK). Cholesterol (total), HDL and LDL cholesterol were analysed on Randox RX Series analysers (RCLS, Antrim, UK). Iron was analysed on RX Imola analysers (Randox Laboratories Ltd, Crumlin, UK). Homocysteine was analysed on Siemens Immulite XP 2000 analysers (Siemens Healthineers AG, Munich, Germany). Human tissue-type plasminogen activator (tPA) and human type 1 plasminogen activator inhibitor PAI-1/tPA complex ELISAs were obtained from AssayPro, (3400 Harry S. Truman Blvd, St. Charles, MO63301). Folate and vitamin Bi2 were analysed on Roche Cobas e801 analysers (Roche Diagnostics, Risch- Rotkreuz, Switzerland). Assays were completed according to manufacturer’s instructions. The limits of detection (LOD) for the biomarkers under investigation were as follows: Cytokine I - IL-2 2.97 pg/ml; IL-4 2.12 pg/ml; IL-6 0.12 pg/ml, IL-8 0.36 pg/ml; VEGF 3.24 pg/ml; IFNy O.44 pg/ml; TNFa 0.59 pg/ml; IL1a 0.19 pg/ml; MCP1 3.53 pg/ml; EGF 1.04 pg/ml; IL-10 0.37 pg/ml; I L-1 b 0.26 pg/ml; Metabolic Array I - Ferritin 3.27 ng/ml, insulin 2.32 mI U/m I , leptin 1.10 ng/ml, PAI-1 2.34 ng/ml, and resistin 1.06 ng/ml; Metabolic Array II - CRP 0.69 mg/I, adiponectin 164 ng/ml, and cystatin C 60 ng/ml; Cerebral Array I - BDNF 0.59 pg/ml, and GFAP 0.18 ng/ml; Cerebral Array II - D-dimer 2.1 ng/ml, NSE 0.26 ng/ml, NGAL 17.8 ng/ml, and STNFRI 0.24 ng/ml. Direct HDL - cholesterol (HDL) 0.189 mmol/l (7.30 mg/dl), direct LDL - cholesterol (LDL) 0.189 mmol/l (7.30 mg/dl), cholesterol 0.865 mmol/l (33.4 mg/dl). Roche Cobas - Folate - 1.2 ng/ml (2.72 nmol/L), Vitamin BI ₂ - 100 pg/ml (73.8 pmol/L). Siemens Immulite - Homocysteine - 1.2 pmol/L. RX Imola - Iron - 2 pmol/L. AssayPro ELISA - Human tPA ELISA - 0.013 ng/ml and human PAI-1 /tPA complex ELISA - 0.05 ng/ml. Thirty-two biomarkers in serum were investigated. Biomarker values below the LOD were recorded as 90% of LOD.

Statistical Analyses

Statistical analyses were performed using the R computer program, Version 3.5.1 (2018: The R Foundation for Statistical Computing)(7). Continuous variables are presented as mean ± standard deviation (mean ± SD). All tests were performed in duplicate. Comparisons were made using the Wilcoxon signed rank test for non- parametric data. Categorical variables are presented as percentage (%) and were compared using a chi square (c ² ) test. A two-tailed p value of <0.05 was considered significant. Least Absolute Shrinkage and Selection Operator (Lasso) regression was used to identify factors in PTSD patients. The Area Under Receiver Operator Curve (AUROC), sensitivity and specificity were calculated for each significant factor (clinical and biomarker) and combined to differentiate between control (non-PTSD) and PTSD participants.

Results

Smoking and Alcohol Consumption

It is well known that anxious people smoke more (8). However, surprisingly, in this study, the control participants smoked more (8.6±9.9 vs. 2.8±7.2 cigarettes/day, p=0.003) and for longer (9.2±9.8 vs. 3.6±8.3 years, p=0.003). Furthermore, there were more participants that smoked in the control vs. PTSD group (24 vs. 12, p=0.006, respectively). Interestingly, alcohol consumption was also significantly higher in the control vs. PTSD group (23 vs. 14, p=0.041 , respectively). However, the incidence of substance abuse (alcohol) was significantly higher amongst PTSD participants (p=0.011).

BMI, Pulse, Hypertension (HTN), Systolic and Diastolic Blood Pressure

There was no significant difference in BMI, pulse rate, HTN, or systolic blood pressure between the control and PTSD participants. However, in agreement with other studies (9), we did observe a significant increase in the diastolic blood pressure for the PTSD participants (p=0.004), suggesting that exposure to trauma may increase blood pressure in this population of individuals.

Pain, Social Support, Comorbidities and Medications Pain

PTSD participants reported a significantly higher incidence of both acute and chronic pain (p=0.020 and p=0.050, respectively). When participants in the study were asked to describe where they experienced pain, they reported: pain in their head, neck, mouth, tongue, shoulder, lower back, hip, arm, knee, fingers, ankle, toes and feet. The average pain (scored from 0 = no pain, 10 = extreme pain), least and worst pain over 24 hours was also significantly higher in the PTSD participants (p=0.017, p=0.007 and p=0.026, respectively). In addition, almost half of the PTSD participants commented that pain interfered with their sleep (8/18 (44.4%) vs. 7/39 (17.9%), p=0.035, respectively). Participants were also asked to describe their pain by selecting one or more of the following conditions: aching, stabbing, sharp, exhausting, nagging, unbearable, squeezing, throbbing, gnawing, tender, tiring, numb, cramping, shooting, burning, penetrating, miserable, radiating and/or deep. Participant results are described in Table 2.

Social Support PTSD not only affects an individual’s mental health, but it can negatively affect others. A spouse can feel isolated, alienated and frustrated from the inability to work through problems. It is noteworthy therefore that there were more individuals that were single in the PTSD vs. control group (64.1% vs. 34.2%, p=0.009, respectively) and more individuals that were married in the control group vs. PTSD participants (55.3% vs. 20.5%, p=0.002, respectively) (Table 3). It is also of interest that the divorce rate was not significantly different between the control and PTSD group (10.5% vs. 12.8%, p=0.754, respectively), suggesting that PTSD may impact the ability to form relationships and not necessarily the ability to maintain them.

Comorbidities PTSD participants presented with significantly more comorbidities than the control group e.g. diabetes, panic disorder, substance abuse, ADD, depression, and anxiety. In addition, over 20% of the PTSD participants reported that they had attempted, at one stage in their life, to commit suicide compared to 0% in the control group (p=0.003).

Medications

As detailed earlier, PTSD participants presented with host of comorbidities, so it is not surprising that these individuals were prescribed significantly more medications when compared to the control group (2.8±2.4 vs. 0.7±1.2, p<0.001 , respectively). Three main categories of medications were prescribed; antidepressants, antianxiety, and pain relief for the following conditions: PTSD (Drug name - Abilify, Ativan, Celexa, Cymbalta, Depakote, Lexapro, Prozac, Prazosin, Seroquel, Trazadone, Zoloft), anxiety (Drug name - Atarax, Bupropin HCI, Buspirone HCI, Clozapine, Hydroxyzine, Klonopin), depression (Drug name - Celex, Duloxetine, Effexor, Fetzima, Lamictal, Latuda, Lexapro, Paroxetine, Seroquel, St. John’s Wort, Venlafaxine, Viibryd, Wellbutrin, Xanax), and pain (Drug name - Baclyfen, Cyclobenzaprine, Gabapentin, Ibuprofen, Lamictal, Meloxicum, Mobic, Norco).

Biomarker Analyses (univariate)

Thirteen biomarkers (15/37 (40.5%)) were identified as being significantly elevated in PTSD participants when compared to the control group; HDL cholesterol, LDL cholesterol, ferritin, folate, homocysteine, adiponectin, IL-8, VEGF, MCP-1 , EGF, D- dimer, NSE, NGAL, BDNF and GFAP (Table 4). Biomarkers identified with the highest predictive value (AUC) for identifying participants with PTSD included NSE (AUC 0.810), HDL cholesterol (AUC 0.791), VEGF (AUC 0.786), folate (AUC 0.785), GFAP (AUC 0.771), D-dimer (AUC 0.768), MCP-1 (AUC 0.757), Ferritin (AUC 0.745), IL-8 (AUC 0.724), BDNF (AUC 0.702), NGAL (0.696), homocysteine (AUC 0.691), LDL cholesterol (AUC 0.673), adiponectin (AUC 0.668) and EGF (AUC 0.639). Biomarkers were selected by Lasso (a regression analyses method that performs both variable selection and regularization to enhance the prediction accuracy and interpretability of the statistical model). Biomarkers selected by Lasso included NSE, HDL cholesterol, VEGF, GFAP and ferritin.

GFAP GFAP is a biomarker for astroglial injury (10) and neurodegeneration (11). In addition, plasma GFAP-BDP levels have been used to identify patients with traumatic brain injury (TBI) and results have been correlated with CT findings (12). Therefore, it was surprising that we detected a significant difference in the level of GFAP in individuals with PTSD vs. the control group (0.9±0.6 vs. 0.4±0.3, p<0.001 , respectively) (Figure 4).

Biomarker Models/Algorithms (multivariate analyses)

Lasso regression identified biomarker combinations that could be used to differentiate between the control and PTSD participants. The goal was to obtain a subset of biomarkers that would minimise prediction error for a quantitative response variable. The following biomarker combinations were identified by Lasso as the best predictors of PTSD: HDL cholesterol + NSE + VEGF + GFAP (AUC 0.955; sensitivity 89.7%, specificity 87.2%) (Table 6; Figure 11). However, as the patients had not fasted prior to venous blood sample collection, HDL cholesterol was removed from the model. Removal of HDL cholesterol from the model and reanalyses identified the following biomarker combinations: Ferritin + NSE + VEGF + GFAP (AUC 0.903; sensitivity 82.1%, specificity 87.2%) (Table 6, Figure 12) and Ferritin + NSE + GFAP (AUC 0.899; sensitivity 79.5%, specificity 87.2%) (Table 6, Figure 13). The individual biomarker and biomarker combinations AUROC are described in Table 6.

Biomarker Models/Algorithms (multivariate analyses) by Gender

Surprisingly, when we examined gender, the significance of the clinical characteristics and biomarker and biomarker combinations were different (Tables 7 and 8). Nine/37 (24.3%) biomarkers were identified for males as significant (HDL cholesterol, ferritin, midkine, cystatin C, IL-2, IL-8, VEGF, D-dimer and NSE) and 16/37 (43.2%) significant biomarkers for females (HDL cholesterol, ferritin, folate, homocysteine, insulin, adiponectin, IL-8, VEGF, TNFa, MCP-1 , EGF, D-dimer, NSE, NGAL, BDNF and GFAP). Male biomarker combinations included HDL cholesterol + ferritin + IL-2 + NSE (AUC 0.984, sensitivity 94.1%, specificity 94.4%) (Figure 14). Removing HDL cholesterol from the model gave an AUC 0.935, sensitivity 94.1%, specificity 83.3%, for ferritin + IL-2 + NSE (Figure 15). In the females, HDL cholesterol + MCP-1 + GFAP gave an AUC 0.957, sensitivity 86.4% and specificity 90.5% (Figure 14). Removing HDL cholesterol, MCP-1 + GFAP gave an AUC 0.909, sensitivity 95.5% and specificity 81% (Figure 17).

Biomarkers for comorbidities of PTSD

Following on from these analyses, we have now identified biomarkers and biomarker models for the comorbidities commonly associated with PTSD e.g. depression, anxiety, panic disorder, substance abuse, and attempted suicide (Table 9 and 10).

Depression

Eleven/32 (34.4%) biomarkers were identified as being significantly elevated in the depression group (n=48 without depression vs. n=24 with depression); HDL cholesterol, adiponectin, VEGF, TNFa, MCP-1 , EGF, D-dimer, NSE, NGAL, BDNF and GFAP. Two/11 biomarkers appeared in the final model for depression; HDL cholesterol and GFAP (AUC 0.882, sensitivity 84.0%, specificity 79.2%).

Anxiety

Ten/32 (31.3%) biomarkers were identified as being significantly elevated in the anxiety group (n=48 no anxiety vs. n=24 with anxiety); Ferritin, HDL cholesterol, insulin, LDL cholesterol, leptin, cystatin C, IL-6, D-dimer, BDNF, and GFAP. Three/9 biomarkers appeared in the final model for anxiety; HDL cholesterol, GFAP and leptin (AUC 0.911 , sensitivity 91.7%, specificity 79.2%). Two/3 of the biomarkers identified for anxiety also appeared in the model for depression suggesting that a potential relationship exists between the two comorbidities. Panic disorder

Six/32 (18.8%) biomarkers were identified as being significantly elevated in the panic disorder group (n=63 controls vs. n=10 panic disorder); CRP, insulin, INFy, I L- 1 a , BDNF and GFAP. Two biomarkers appeared in the final model for panic disorder; GFAP and IL-1a (AUC 0.878, sensitivity 80.0%, specificity 93.7%).

Substance abuse

Eight/32 (25%) biomarkers were identified as being significantly elevated in the substance abuse (alcohol) group (n=65 controls vs. n=6 substance abuse); HDL cholesterol, tPA, PAI-1/tPA, total cholesterol, IL-4, IL-8, IL-10, and TNFa. Three biomarkers appeared in the final model for substance abuse; HDL cholesterol, PAI- 1/tPA and IL-4 (AUC 0.897, sensitivity 83.3%, specificity 84.6%). Suicide attempt

Eight/32 (18.8%) biomarkers were identified as being significantly elevated in the participants who had attempted to take their own life (n=66 non-suicide attempt participants vs. n=7 who had attempted suicide); Ferritin, insulin, tPA, PAI-1/tPA, IL- 10, IL-1 a, I L-1 b and GFAP. However, only 1/8 of the biomarkers appeared in the final model for attempted suicide; ferritin (AUC 0.738, sensitivity 57.1%, specificity 83.3%).

Table 1 Socio-demographic and clinical factors of the study participants Ί 0 = no pain; 10 = extreme pain

Continuous variables were expressed as mean ± SD or as % for categorical variables. The difference in continuous variables were analysed using Wilcoxon signed rank test, while chi square (c ² ) test was used for categorical variables. A p < 0.05 was considered significant.

Table 2 Type of pain identified by study participants

NB: study participants may have selected more than one type of pain when completing questionnaire

Table 3 Marital status of the study participants

Table 4 Serum biomarker levels in control (n=39) vs. clinically-diagnosed PTSD (n=39) study participants assigned 0 and any value above the LOD assigned 1. Chi square (c2) test was used for this categorical variable. Continuous variables were expressed as mean ± SD (Wilcoxon test). Values below the LOD were assigned a value of 90% of the LOD. A p < 0.05 was considered significant. Table 5 Serum biomarkers in controls vs. any mental disorder

Continuous variables were expressed as mean ± SD (Wilcoxon test). Values below the LOD were assigned a value of 90% of the LOD. A p < 0.05 was considered significant.

Table 6 Area Under the Receiver Operator Curve (AUROC), sensitivity and specificity, PPV and NPV for the biomarker and biomarker combinations (algorithm/model) used to differentiate between control and PTSD participants (all genders)

Table 7 Clinical characteristics of the study participants by gender

Table 8 Biomarker and biomarker combinations for PTSD by gender

Table 9 Comorbidities associated with post-traumatic stress disorder Table 10 Serum biomarker models identified by linear regression as diagnostic for comorbidities associated with post-traumatic stress disorder

Table 11 Serum biomarker models identified by linear regression and single markers as diagnostic for any mental disorder vs. controls

PCL-5 control group analysis

PCL-5 scores were only available for 20 of the control participants so further analyses were carried out on an age and sex-matched subset of these 20 controls and 20 PTSD participants. Participant demographics

There was no significant difference in age between the control and PTSD participants (39.0 + 2.6 vs. 41.5 + 11.0, p=0.386, respectively) or male gender (50% vs. 45%, p=0.752). However, a significant difference in ethnicity was observed; a greater proportion of study participants in the PTSD group were non-Caucasian vs. the control group (60% vs. 21.2%, p=0.013, respectively). Comparison of PTSD and control group demographics are further detailed in Table 12.

BMI, Pulse, Hypertension (HTN), Systolic and Diastolic Blood Pressure

There was no significant difference in BMI, pulse rate, HTN, or systolic blood pressure between the control and PTSD participants. However, a significant increase in diastolic blood pressure was observed for PTSD participants (79.9±7.3 vs. 93.6±13.7; p=0.010) (Table 12).

Comorbidities

PTSD participants presented with significantly more comorbidities than the control group i.e. panic disorder, substance abuse, ADD, depression, and anxiety. In addition, 10% of the PTSD participants reported that they had attempted, at one stage in their life, to commit suicide compared to 0% in the control group (Table 12).

Medications

As detailed earlier, PTSD participants presented with more comorbidities, therefore is was unsurprising that PTSD participants were prescribed significantly more medications when compared to the control group (3.2±2.8 vs. 1.1 ±1.4; p=0.008, respectively). Three main categories of medications prescribed included; anti depressants, anti-anxiety, and pain management drugs.

Smoking and Alcohol Consumption

In this study, the control participants smoked more (6.2±7.6 vs. 2.3±4.7 cigarettes/day, p=0.048) and for longer periods of time (8.0±7.9 vs. 3.2±6.3 years, p=0.025). Furthermore, there were more participants that smoked in the control vs. PTSD group (13/20 (65%) vs. 6/20 (30%); p=0.027, respectively). Alcohol consumption was also significantly greater in the control vs. PTSD group (15/20 (75%) vs. 5/20 (25%); p=0.002, respectively). However, the incidence of substance abuse (alcohol) was significantly greater amongst PTSD participants (4/20 (25%) vs. 0/20 (0%); p=0.035) (Table 12).

Social Support Examination of information available on social support networks identified a greater proportion of the PTSD group were single (15/20 (75%) vs. 3/19 (15.8%); p<0.001 , respectively), and fewer were married compared to the control group (4/20 (20.0%) vs. 12/19 (63.2%), p=.006, respectively). The divorce rate was not significantly different between the control or PTSD group (4/19 (21.1%) vs. 1/20 (5.0%); p=0.134, respectively).

Table 12 Socio-demographic and clinical factors of the study participants

_ Factor Control (n=20) PTSD (n=20) p value

Age (years) _ 39.0 ± 2.6 41.5 ± 11.0 0.386

Gender (Male) _ 10/20 (50%) 9/20 (45%) 0.752

Caucasian (Ethnicity) 15/19 (78.9%) 8/20 (40%) 0.013 Smoker _ 13/20 (65%) 6/20 (30%) 0.027

Cigarettes/day _ 6.2 ± 7.6 2.3 ± 4.7 0.048

Smoking Years 8.0 ± 7.9 3.2 ± 6.3 0.025 Alcohol _ 15/20 (75%) 5/20 (25%) 0.002

BMI 29.7 ± 7.9 27.9 ± 6.3 0.496

Pulse _ 73.2 ± 10.4 78.3 ± 19.7 0.725

Systolic BP _ 122.2 ± 14.7 133.8 ± 19.7 0.055

Diastolic BP 79.9 ± 7.3 93.6 ± 13.7 0.010

Diabetes 0/20 (0%) 3/20 (15%) 0.072

Hypertension 4/20 (20%) 4/20 (20%) 1.000

Panic Disorder 0/20 (0%) 5/20 (25%) 0.017

Substance Abuse 0/20 (0%) 4/20 (20%) 0.035

ADD 0/20 (0%) 1/20 (5%) 0.311

Bipolar Disorder 0/20 (0%) 2/20 (10%) 0.147

Depression 4/20 (20%) 13/20 (65%) 0.004

Suicide attempt 0/20 (0%) 2/20 (10%) 0.147

Anxiety 2/20 (10%) 16/20 (80%) <0.001

Medications 1.1 ± 1.4 3.2 ± 2.8 0.008 Continuous variables were expressed as mean ± SD or as % for categorical variables. The difference in continuous variables were analysed using Wilcoxon rank sum test, while chi square (c2) test was used for categorical variables. A p < 0.05 was considered significant. PTSD assessment and symptom severity Total symptom severity for PTSD was determined by summing the severity of each DSM-5 PTSD symptom cluster. Mean total severity of PTSD symptoms for the PTSD group was 47.7 + 7.6. In the control group, mean total severity of PTSD symptoms, as assessed by PCL-5 was 10.1 + 7.8. Severity of individual symptom clusters are shown in Table 13.

Table 13 Summary of the PTSD scale for DSM-5 (CAPS-5) and PCL-5 score

Control PTSD

P Control PTSD å (PCL-5) CAPS-5 value (median) (median)

(n=20) (n=20)

BCDE total 10.1 ± 7.8 47.7 ± 7.6 <0.001 7.0 45.0

BCDE

1.8 ± 2.2 16.6 ± 1.6 <0.001 1.0 17.0

Severity

B total 2.1 ± 2.2 13.3 ± 2.4 <0.001 1.5 13.0

C total 1.0 ± 1.2 5.8 ± 1.0 <0.001 0.5 6.0

D total 2.8 ± 3.0 16.6 ± 3.1 <0.001 2.0 16.0

E total 4.2 ± 3.1 11.9 ± 3.2 <0.001 4.0 11.0

Threshold for CAPS-5/PCL-5 indication of PTSD total score åBCDE >33: B total - intrusion symptoms; C total - avoidance symptoms; D total - cognitions and mood symptoms; E total - arousal and reactivity symptoms

Biomarker Analysis

In total, 37 serum biomarkers were investigated (Table 14). Twelve biomarkers (12/37 (32.4%)) were identified as significantly altered in PTSD participants when compared to the control group; Biomarkers - HDL cholesterol, LDL cholesterol, EGF, tPA, IL-8, PAI-1 , D-dimer, cystatin C, homocysteine, midkine, folic acid, and IFNy (Table 14). LDL and HDL cholesterol had the highest AUC for differentiating between the control vs. PTSD group. However, as the study participants had not been asked to fast prior to their venous blood sampling, HDL and LDL cholesterol were not included in any further analyses. Of the remaining 10 significant biomarkers, 5 were significantly lower in the PTSD group; EGF (13.3 + 14.3 pg/ml vs. 73.2 + 35.7 pg/ml, p <0.001), PAI-1 (24.0 + 8.1 ng/ml vs. 35.0 + 12.0 ng/ml, p=0.003), D-dimer (28.1 + 27.9 ng/ml vs. 50.1 + 38.6 ng/ml, p=0.006), Cystatin C (1.0 + 0.2 pg/ml vs. 1.2 + 0.2 pg/ml, p =0.007), and Homocysteine (8.9 + 2.8 pmol/l vs. 12.5 + 7.4 pmol/l, p =0.014)(Figure. 1). Biomarkers that were significantly elevated in PTSD included; tPA (5.1 + 2.9 ng/ml vs. 1.7 + 0.7 ng/ml, p<0.001), IL-8 (13.0 + 5.6 pg/ml vs. 8.1 + 2.8 pg/ml, p<0.001), Midkine (482.4 + 979.9 pg/ml vs. 107.0 + 42.2 pg/ml, p =0.023) and Folic Acid (15.1 + 5.9 pg/ml vs. 12.7 + 8.5 pg/ml, p=0.048) (Figure. 2). For lNFy, results were converted to nominal data. Values below the LOD for INFy were assigned 0, and those above the LOD were assigned 1. A significantly greater proportion of the PTSD group had INFy results above the LOD, compared to controls

(50% vs. 10%, p=0.006). The sensitivity and specificity for how each marker can distinguish PTSD and control participants is summarised by their Area Under the Receiver Operator Curve (AUROC) and presented in Table 14.

Table 14 Comparison of mean serum biomarker levels in control vs. PTSD participants. Area Under the Receiver Operator Curve (AUROC), sensitivity and specificity, positive predictive value (PPV) and negative predictive value (NPV) are shown for each biomarker Adiponectin (pg/ml) 4.1 ± 4.0 3.9 ± 2.3 0.314 0.595 0.700 0.600 63.6 66.7

IL1B (pg/ml) 1.8 ± 0.3 2.2 ± 1.0 0.317 0.594 0.650 0.550 59.1 61.1

Leptin (mg/I) 12.9 ± 13.9 12.3 ± 14.1 0.337 0.593 0.450 0.900 81.8 62.1

IL-4 (pg/ml) 2.6 ± 0.4 2.4 ± 0.4 0.309 0.589 0.550 0.650 61.1 59.1

IL1A (pg/ml) 0.5 ± 0.1 0.4 ± 0.1 0.364 0.584 0.450 0.750 64.3 57.7

MCP1 (pg/ml) 188.8 ± 117.3 186.2 ± 71 .7 0.441 0.573 0.650 0.550 59.1 61.1

STNFR1 (ng/ml) 0.8 ± 0.2 0.8 ± 0.2 0.441 0.573 0.400 0.750 61.5 55.6

IL-2 (pg/ml) 3.3 ± 0.8 3.5 ± 1.0 0.485 0.553 0.600 0.600 60.0 60.0

Iron (pmol/l) 12.2 ± 5.2 13.2 ± 9.7 0.646 0.544 0.600 0.700 66.7 63.6

IL-6 (pg/ml) 2.4 ± 3.0 2.7 ± 2.8 0.655 0.543 0.600 0.550 57.1 57.9

IL-10 (pg/ml) 1.3 ± 0.2 1.7 ± 1.3 0.715 0.535 0.500 0.700 62.5 58.3

BDNF (pg/ml) 7099.5 ± 743.9 6407.7 ± 1704.1 0.328 0.533 0.600 0.650 63.2 61.9

VEGF (pg/ml) 80.5 ± 52.6 95.1 ± 94.8 0.776 0.528 0.700 0.500 58.3 62.5

FABP (ng/ml) 1.4 ± 0.5 1.4 ± 0.7 1.000 0.501 0.450 0.750 64.3 57.7

NSE (ng/ml) 4.0 ± 1.6 4.8 ± 3.3 1.000 0.499 0.400 0.750 61.5 55.6

Total Cholesterol (mmol/l) 5.1 ± 1.0 5.1 ± 1.1 0.839 0.480 0.550 0.550 55.0 55.0

TNFa (pg/ml) 2.2 ± 0.4 2.4 ± 0.8 0.829 0.479 0.500 0.600 55.6 54.5

* IFNy results were converted to nominal data with values below the LOD assigned 0 and any value above the LOD assigned 1 . Chi square (c2) test was used for this categorical variable. Continuous variables were expressed as mean ± SD (Wilcoxon rank sum test). Values below the LOD were assigned a value of 90% of the LOD. A p<0.05 was considered significant. LDL and HDL markers are reported but excluded from analysis as participants had not been asked to fast prior to venous blood sample collection.

Biomarker Combinations (algorithms)

Lasso regression identified biomarker combinations that could be used to differentiate between the control and the PTSD participants. The goal was to obtain a subset of biomarkers that would minimise prediction error for a quantitative response variable (Table 15). Resultant biomarker combinations are displayed in Table 15. Bivariate biomarker combinations with an AUC value above 0.9 include; tPA + Cystatic C (AUC 1.00); EGF + tPA (AUC 0.985); tPA + PAI-1 (AUC 0.978), and PAI-1 + IL-8 (AUC 0.950). Further biomarker algorithms identified by Lasso are shown in Table 15.

Table 15 Area Under the Receiver Operator Curve (AUROC), sensitivity and specificity, positive predictive value (PPV) and negative predictive value (NPV) for biomarker models used to differentiate between the control and PTSD participants

AUR Sensitiv Specific PPV NPV

Model Biomarker Models OC ity ity (%) (%)

A _ tPA + Cystatin C _ 1.000 1.000 1.000 100.0 100.0

B EGF+ tPA 0.985 1.000 0.900 90.9 100.0 C tPA + PAI-1 0.978 0.900 1.000 100.0 90.9

D PAI-1 + IL-8 0.950 0.900 0.900 90.0 90.0

E PAI-1 + Cystatin C 0.850 0.850 0.850 85.0 85.0

F IL-4 + IL-8 0.833 0.800 0.800 80.0 80.0

G PAI-1 + IL-4 0.778 0.850 0.600 68.0 80.0

H tPA + PAI-1 + Cystatin C 1.000 1.000 1.000 100.0 100.0

I PAI-1 + IL-4 + IL-8 0.950 0.900 0.900 90.0 90.0

J IFNy + IL-8 + PAI-1 -tPA + tPA + PAI-1 + Cystatin C 1.000 1.000 1.000 100.0 100.0

K EGF+ IFNy+- IL-8 + PAI-1 - tPA + tPA + PAI-1 + Cystatin C 1.000 1.000 1.000 100.0 100.0