Lipid profile disturbances in antipsychotic-naive patients with first ...

SCHRES-07224; No of Pages 10 Schizophrenia Research xxx (2017) xxx–xxx

Contents lists available at ScienceDirect

Schizophrenia Research journal homepage: www.elsevier.com/locate/schres

Lipid profile disturbances in antipsychotic-naive patients with first-episode non-affective psychosis: A systematic review and meta-analysis Błażej Misiak a,⁎, Bartłomiej Stańczykiewicz b, Łukasz Łaczmański c, Dorota Frydecka b a b c

Department of Genetics, Wroclaw Medical University, 1 Marcinkowski Street, 50-368 Wroclaw, Poland Department of Psychiatry, Wroclaw Medical University, 10 Pasteur Street, 50-367 Wroclaw, Poland Ludwik Hirszfeld Institute of Immunology and Experimental Therapy of the Polish Academy of Sciences, 12 Weigla Street, 11-400 Wroclaw, Poland

a r t i c l e

i n f o

Article history: Received 23 January 2017 Received in revised form 10 March 2017 Accepted 11 March 2017 Available online xxxx Keywords: Schizophrenia Lipids Lipoproteins Metabolic dysregulation Cholesterol Psychotic disorders

a b s t r a c t Background: Dyslipidaemia is one of the most prevalent metabolic disturbances observed in schizophrenia patients and has been largely attributed to the effects of poor lifestyle habits and adverse effects of antipsychotic treatment. However, less is known whether patients with first-episode non-affective psychosis (FENP) present subthreshold indices of dyslipidaemia. Therefore, we tested the hypothesis whether subclinical lipid profile alterations occur already in antipsychotic-naïve FENP patients. Methods: In this systematic review and meta-analysis we adhered to the PRISMA guidelines and searched PubMed, CINAHL Complete, Academic Search Complete, ERIC and Health Source: Nursing/Academic Edition from database inception to Dec 12, 2016, for case-control studies measuring the levels of total cholesterol, low- and high-density lipoproteins (LDL and HDL) and triglycerides in patients with FENP and controls. W calculated effect size (ES) estimates as Hedges' g and pooled data using random- or fixed-effects models depending on heterogeneity. Our study was registered in the PROSPERO database (CRD42016051732). Results: Out of 2466 records identified, 19 studies representing 1803 participants were finally included in our systematic review and meta-analysis. Pooled analysis revealed that FENP patients had significantly lower levels of total cholesterol [ES = −0.16 (95% CI: −0.27, −0.06), p = 0.003], LDL [ES = −0.13 (95% CI: −0.24, −0.01), p = 0.034] and HDL [ES = −0.27 (95% CI: −0.49, −0.05), p = 0.018] as well as significantly higher levels of triglycerides [ES = 0.22 (95% CI: 0.11, 0.32), p b 0.001] compared to controls. After removing single studies in sensitivity analysis, ES estimate for LDL levels was insignificant. Conclusions: Antipsychotic-naïve patients with FENP present subclinical dyslipidaemia. Future studies should disentangle whether our findings reflect disease-specific mechanisms. © 2017 Elsevier B.V. All rights reserved.

1. Introduction Cardiovascular morbidity is increasingly being recognized as a major contributor of mortality and reduced life expectancy in patients with schizophrenia (Laursen et al., 2014). Indeed, this group of patients presents with higher prevalence of metabolic syndrome together with its single components, and related abnormalities, including hyperglycaemia, obesity, lipid profile disturbances, elevated homocysteine levels and cigarette smoking (de Leon and Diaz, 2005; Mitchell et al., 2013; Muntjewerff et al., 2006; Vancampfort et al., 2015). Interestingly, a meta-analysis by Mitchell et al. (2013) revealed that about 40% of patients with schizophrenia and related disorders have lipid profile ⁎ Corresponding author at: Department of Genetics, Wroclaw Medical University, 1 Marcinkowski Street, 50-368 Wroclaw, Poland. E-mail address: [email protected] (B. Misiak).

alterations, which might be the most prevalent cardiovascular risk factor in this population. High cardiovascular risk is largely attributable to environmental factors related to unhealthy lifestyle characteristics and the effects of antipsychotic treatment. In addition, patients with schizophrenia do not receive adequate monitoring of cardiovascular health and tackle with inequalities in the availability of healthcare provision (Baller et al., 2015; Lawrence and Kisely, 2010). Accumulating evidence indicates that subthreshold metabolic dysregulation might be present already in the premorbid phase of the illness and in antipsychotic-naïve patients with first-episode psychosis. Indeed, there are studies showing that subjects at risk of psychosis have higher rates of metabolic syndrome and reduced high density lipoproteins (HDL) levels as well as higher levels of fasting blood glucose and blood pressure (Cordes et al., 2016). These abnormalities might be attributed to low levels of physical activity and high rates of cigarette smoking or alcohol abuse (Carney et al., 2016). In turn, studies

http://dx.doi.org/10.1016/j.schres.2017.03.031 0920-9964/© 2017 Elsevier B.V. All rights reserved.

Please cite this article as: Misiak, B., et al., Lipid profile disturbances in antipsychotic-naive patients with first-episode non-affective psychosis: A systematic review and met..., Schizophr. Res. (2017), http://dx.doi.org/10.1016/j.schres.2017.03.031

2

B. Misiak et al. / Schizophrenia Research xxx (2017) xxx–xxx

performed on antipsychotic-naïve first-episode psychosis (FENP) patients indicate abnormal glycaemic control (Greenhalgh et al., 2016; Perry et al., 2016) as well as elevated levels of C-reactive protein and pro-inflammatory cytokines (Fernandes et al., 2016; Upthegrove et al., 2014). Early metabolic dysregulation observed in antipsychotic-naïve patients with FENP suggests that schizophrenia-spectrum disorders might share overlapping genetic background with cardio-metabolic phenotypes. A recent analysis of combined data from genome-wide association studies (GWASs) identified ten loci associated with both schizophrenia and cardiovascular risk factors, mainly the levels of low- and high-density lipoproteins (LDL and HDL), but also waist-tohip ratio, systolic blood pressure, and body-mass index (BMI) (Andreassen et al., 2013). Accumulating evidence also indicates that schizophrenia might be associated with alterations in biosynthesis of cholesterol, fatty acids, phospholipids and sphingolipids (Schneider et al., 2016; Steen et al., 2016). Indeed, there are studies showing alterations in the levels of polyunsaturated fatty acids, cholesteryl esters, phospholipids and triglycerides in prefrontal and frontal cortex of patients with schizophrenia (Horrobin et al., 1991; McNamara et al., 2007; Taha et al., 2013). Finally, it has been widely reported that patients with schizophrenia have various myelination abnormalities that might be attributed to genetic factors (Mighdoll et al., 2015). In turn, previous studies addressing peripheral blood lipid profile alterations in the early course of psychosis have provided mixed results. A recent meta-analysis by Perry et al. (2016) revealed decreased levels of total cholesterol and HDL. However, this meta-analysis primarily investigated parameters of glucose homeostasis and included patients with affective and non-affective psychosis. Therefore, in this study we performed a systematic review and meta-analysis of lipid profile alterations in antipsychotic-naïve patients with FENP. 2. Methods and materials 2.1. Search strategy Independent online search was performed by two authors (BS and BM), using the following combination of single keywords from the following groups: 1) “antipsychotic-naïve”, “antipsychotic-free” “drugnaïve”, “drug-free”, “neuroleptic-naïve”, “neuroleptic-free”, “nevermedicated” and “untreated”; 2) “first-episode psychosis”, “first-episode schizophrenia”, “FEP”, “FES”, “psychosis”, and “schizophrenia” and 3) “cholesterol”, “HDL”, “LDL”, “triglycerides”, “lipid”, “lipoprotein”, and “metabolic”. Relevant publications were identified in the following databases: PubMed, CINAHL Complete, Academic Search Complete, ERIC and Health Source: Nursing/Academic Edition. Additionally, our online search was supplemented by reference lists from relevant publications. We included studies that determined peripheral blood levels of total cholesterol, HDL, LDL and triglycerides in adult antipsychotic-naïve patients with FENP (DSM-IV and ICD-10 criteria) and healthy controls. Studies were included if necessary data was available in the article or upon request (contact with corresponding authors). Our exclusion criteria were: 1) non-English language publications, 2) non-original publications (commentaries, editorials, hypotheses, reviews and study protocols) and meta-analyses, 3) animal model studies and 4) studies that did not include a group of matched healthy controls. Discrepancies were resolved through discussion with the third author (DF). Search strategy followed PRISMA guidelines (Moher et al., 2009). 2.2. Data analysis The following data of antipsychotic-naïve FENP patients and healthy controls were extracted from eligible publications by one author (BM): 1) age; 2) body-mass index (BMI); 2) the levels of total cholesterol, HDL, LDL and triglycerides; 3) the number of male and female participants and 4) the number of cigarette smokers. Continuous variables extracted

from relevant publications were expressed as mean ± SD. In case of potential duplicate data, corresponding authors of relevant publications were asked to provide data from the whole sample of participants without duplicates. Quality of studies was assessed using the Newcastle-Ottawa Scale (NOS) for case-control studies (Wells et al., 2000). Publication bias was analysed using funnel plots of asymmetry, the Begg & Mazumbar's test and the Egger's test. Effect size (ES) estimates were calculated as Hedges' g with corresponding 95% confidence intervals (95% CI). Fixed-effects models were used unless significant heterogeneity was demonstrated. Otherwise, all analyses were performed under random-effects models. Heterogeneity between studies was evaluated using the Cochran's Q test and I2 estimates. Sensitivity analysis was performed in two steps: 1) by removing one study at a time and repeating calculations to examine the impact on the ES estimates; 2) by removing studies with significant differences between patients and controls in at least one of potential confounding variables, such as age, gender distribution, BMI or cigarette smoking, together with outliers, and repeating calculations to examine the impact on the ES estimates. Following the second step of sensitivity, studies were dichotomized as those with perfect matching and those with at least one significant difference in potential confounding variables. This dichotomous variable and the NOS score were included in meta-regression. Results were considered as statistically significant if the p-value was b0.05. Statistical analysis was performed using the STATISTICA software, version 12.5. Our systematic review with meta-analysis was registered in the PROSPERO database on 28th November 2016 (registration number: CRD42016051732). 3. Results Out of 2466 records identified, 57 full-text articles were assessed for eligibility and 19 studies met our inclusion and exclusion criteria (Fig. 1). There were 15 case-control studies (McCreadie and the Scottish Schizophrenia Research Group, 2000; Chen et al., 2016; Dasgupta et al., 2010; Enez Darcin et al., 2015; Hepgul et al., 2012; Kavzoglu and Hariri, 2013; Kirkpatrick et al., 2010; Misiak et al., 2016a; Petrikis et al., 2015; Ryan et al., 2003; Sengupta et al., 2008; Spelman et al., 2007; Venkatasubramanian et al., 2007; Wu et al., 2013; Zhang et al., 2016) and 4 observational studies with baseline comparisons of FENP patients and healthy controls (Basoglu et al., 2010; Cai et al., 2012; Nyboe et al., 2015; Saddichha et al., 2008) (Table 1). In total, 1803 participants were recruited in studies included in this meta-analysis (866 FENP patients and 937 controls with mean age of 28.0 years and 28.7 years, respectively). Sample size of antipsychotic-naïve FENP patients varied from 3 (Hepgul et al., 2012) to 172 (Chen et al., 2016) cases, while the number of controls was between 11 (Cai et al., 2012) to 146 (Misiak et al., 2016a) participants. There was a relative predominance of males in the group of FENP patients (58.4%) and controls (55.2%). Patients and controls were matched for age and sex in almost all studies with exception of two studies (Enez Darcin et al., 2015; Zhang et al., 2016). There were significantly more males in the group of patients compared to the group of controls in the study by Enez Darcin et al. (2015). In turn, in the study by Zhang et al. (2016), patients were significantly younger than controls. In the majority of studies, authors measured BMI (except for one study (Saddichha et al., 2008)) and recorded information about cigarette smoking status (McCreadie and the Scottish Schizophrenia Research Group, 2000; Basoglu et al., 2010; Cai et al., 2012; Chen et al., 2016; Dasgupta et al., 2010; Enez Darcin et al., 2015; Kavzoglu and Hariri, 2013; Misiak et al., 2016a; Nyboe et al., 2015; Petrikis et al., 2015; Ryan et al., 2003; Spelman et al., 2007; Zhang et al., 2016). In two studies, BMI was significantly lower (McCreadie and the Scottish Schizophrenia Research Group, 2000; Spelman et al., 2007), while in one study it was significantly higher (Nyboe et al., 2015) in the group of patients compared to controls. In one study (Spelman et al., 2007), the number of cigarette smokers


3

Fig. 1. Selection of studies for systematic review and meta-analysis.



4

Study

FENP patients

Controls

Methodology

n (M/F)

Age (mean ± SD), years n (M/F)

Age (mean ± SD), years

Chen et al. (2016)

172 (83/89)

28.7 ± 9.9

31 (14/17)

26.9 ± 5.2

Misiak et al. (2016a)

24 (14/10)

26.8 ± 2.9

146 (67/79)

27.6 ± 5.5

Zhang et al. (2016)

31 (15/16)

25.6 ± 5.1

71 (32/39)

30.2 ± 5.0

Enez Darcin et al. (2015)

40 (29/11)

34.6 ± 1.1

70 (37/33)

34.6 ± 8.9

Nyboe et al. (2015)

17 (8/9)a

22.4 ± 3.1

50 (29/21)

23.1 ± 4.6

Petrikis et al. (2015)

40 (25/15)

31.9 ± 8.3

40 (27/13)

32.4 ± 9.8

Wu et al. (2013)

70 (37/33)

24.5 ± 7.0

44 (20/24)

26.2 ± 4.2

Kavzoglu and Hariri (2013)

50 (25/25)

30.1 ± 7.5

50 (21/29)

30.9 ± 6.4

Case control study: FES patients recruited from consecutive inpatient admissions. A diagnosis of schizophrenia was validated using SCID (assessment at baseline and confirmed after 3 months). Controls were recruited through advertisement in the local area (assessment with unstructured interviews). Controls were excluded in case of any medical illness and drug and/or alcohol abuse/dependence (with exception of cigarette smoking). Both groups were matched for age, sex, education, BMI, waist circumference, WHR and cigarette smoking. Lipid profile parameters: TC, LDL, HDL and TG. Case control study: FES patients recruited from consecutive inpatient admissions. A diagnosis of schizophrenia was validated using OPCRIT. Controls were recruited from hospital staff, students and by the word of mouth (assessment with unstructured interviews). They had no comorbid medical illnesses and substance and/or alcohol dependence. Both groups were matched for age, sex, education, BMI, waist circumference and cigarette smoking. Lipid profile parameters: TC, LDL, HDL and TG. Case control study: FES patients with schizophrenia diagnosed in accordance with DSM-IV criteria. No information about the way controls were recruited. Participants were excluded in case of physical and metabolic illnesses. Patients were significantly younger than controls. Both groups were matched for sex, BMI and cigarette smoking. Lipid profile parameters: TC, LDL, HDL and TG. Case control study: FES patients diagnosed in accordance with DSM-IV criteria out of 1241 patients with schizophrenia admitted in the study period. Controls were recruited from relatives of the clinical staff and had no physical illnesses. Patients and controls were matched for age and BMI. Educational attainment was significantly lower in the group of patients. There were significantly more males in the group of patients. Lipid profile parameters: TC, LDL, HDL and TG. 1-year follow-up study: FES patients diagnosed in accordance with ICD-10 criteria. Controls recruited via advertisement in the local newspaper. Participants with physical disability or somatic illness impairing physical activity as well as substance or alcohol abuse (ICD-10) were excluded. Patients and controls were matched for age and sex. BMI was significantly higher in the group of patients. Educational attainment was significantly lower in the patients' group. Lipid profile parameters: TC, LDL, HDL and TG. Case control study: Patients with schizophrenia, schizophreniform disorder and brief psychotic episode, diagnosed using SCID (patients with comorbid substance and or alcohol abuse/dependence were excluded). Controls were recruited through advertising amongst university students and employees as well as from local enterprise employees. All control subjects were physically healthy and were not taking any medications. They were excluded in case of substance and/or alcohol abuse. Both groups were matched for age, sex, BMI and cigarette smoking. Lipid profile parameters: TC, HDL and TG. Case control study: FES patients were recruited and psychiatric diagnosis was validated using SCID. A diagnosis of schizophrenia was confirmed 6 months after discharge. Patients with comorbid alcohol/substance abuse/ dependence were excluded. Controls were recruited through advertising from university students, the same hospital employees and local enterprise employees. All controls were physically healthy and had no DSM-IV diagnosis of substance and/or alcohol abuse. Both groups were matched for age, sex, waist circumference and WHR. Lipid profile parameters: TC, LDL, HDL and TG. Case control study: A diagnosis of FES was established based on DSM-IV criteria. Relatives or friends of the patients were recruited as controls. They had no physical or psychiatric illnesses. Both groups were matched for age, sex, cigarette smoking and BMI. Lipid profile parameters: TC, LDL, HDL and TG.



Table 1 Characteristics of studies included in systematic review and meta-analysis.

11 (6/5)

27.6 ± 9.5

11 (6/5)

27.6 ± 9.5

Hepgul et al. (2012)

3 (2/1)

24.7 ± 5.9

49 (36/13)

26.3 ± 0.6

Basoglu et al. (2010)

27 (27/0)

21.2 ± 0.7

22 (22/0)

21.7 ± 1.1

Dasgupta et al. (2010)

30 (14/16)

32.5 ± 10.5

25 (12/13)

35.7 ± 9.6

Kirkpatrick et al. (2010)

76 (49/27)

27.1 ± 5.3

76 (49/27)

27.0 ± 4.5

Saddichha et al. (2008)

99 (52/47)

26.0 ± 5.5

51 (30/21)

27.5 ± 5.9

Sengupta et al. (2008)

38 (33/5)

25.4 ± 5.6

36 (28/8)

25.1 ± 5.3

Spelman et al. (2007)

38 (28/10)

25.2 ± 5.6

38 (28/10)

25.2 ± 5.7

Venkatasubramanian et al. (2007)

44 (23/21)

33.0 ± 7.7

44 (23/21)

32.5 ± 7.6

Ryan et al. (2003)

26 (15/11)

33.6 ± 13.5

26 (15/11)

34.4 ± 1.9

McCreadie and the Scottish Schizophrenia Research Group (2000)

30 (21/9)

30.5 ± 9.0

30 (21/9)

30.0 ± 7.0

A 6-week follow-up study: Patients with FES and without substance abuse were recruited (a diagnosis based on DSM-IV criteria). Controls were screened for serious medical illnesses and drug/alcohol abuse. Both groups were matched with controls for age, sex, BMI and cigarette smoking status. Lipid profile parameters: TC, LDL, HDL and TG. Case control study: Patients included in this meta-analysis had a diagnosis of non-affective psychosis based on ICD-10 criteria (F20–F29), although authors also assessed patients with affective FEP (F30–F39). Controls were recruited from the same catchment area through advertisements and from database of volunteers. No information about physical health of patients and controls was provided. Both groups were matched for age and sex. Lipid profile parameters: TC, LDL, HDL and TG. A 6-week follow-up study: FES patients referred from the mandatory military service (diagnosis was based on DSM-IV criteria). Healthy volunteers were also recruited from the military service. Controls had no history of head trauma and neurological diseases but information about other physical health impairments was not provided. Both groups were matched for age, sex, BMI and cigarette smoking status. Lipid profile parameters: LDL, HDL and TG. Case control study: Patients with FES and without substance and alcohol abuse (including nicotine) during previous 6 months were recruited (diagnosis was based on DSM-IV criteria). Controls were selected from persons accompanying the patients. Participants with any somatic illness and alcohol/substance abuse/ dependence were excluded. Both groups were matched for age, sex and BMI. Lipid profile parameters: TC, LDL, HDL and TG. Case control study: Patients with FENP according to DSM-IV criteria were enrolled (validation of diagnosis using SCID). Controls were recruited through advertisements. They had no serious medical conditions and diabetes. Both groups were matched for age, sex, BMI and cigarette smoking. Lipid profile parameters: TC, LDL, HDL and TG. A 6-week follow-up study: All consecutively admitted FES patients (DSM-IV criteria) were asked to participate in this study. They did not have comorbid substance abuse or dependence. No information about physical health impairments was provided. FES patients and controls were matched for age, sex, waist circumference, exercise activity and diet. Lipid profile parameters: HDL and TG. Case control study: All patients had schizophrenia-spectrum diagnoses (schizophrenia, schizophreniform disorder, schizoaffective disorder, delusional disorder and non-affective psychotic disorder NOS) based on SCID. Controls were recruited through advertising, from local universities and hospital employees. Participants with major physical illnesses were excluded. Both groups were matched for age, sex, BMI and waist circumference. Smoking rates and WHR were significantly higher in the group of patients. Lipid profile parameters: TC, LDL, HDL and TG. Case control study: FES patients were recruited and diagnosed using SCID. Controls were matched for age, sex, smoking and alcohol intake. They had no physical health impairments. Patients had significantly lower BMI. Lipid profile parameters: TC, LDL, HDL and TG. Case control study: Patients with a DSM-IV diagnosis of FES were recruited (MINI was used to validate a diagnosis). Healthy controls were recruited from the medical staff and their friends. Participants had no physical disorders. Both groups were matched for age, sex and BMI. Lipid profile parameters: TC and TG. Case control study: Patients with a DSM-IV diagnosis of FES were recruited. Controls were recruited within the hospital, the affiliated university and the general community. All participants were physically healthy. Both groups were matched for age, sex, BMI, waist circumference, WHR, diet, exercise activity and cigarette smoking. Lipid profile parameters: TC, LDL, HDL and TG. Case control study: Patients had a DSM-IV diagnosis of FES. Patients and controls were matched for age, sex, cigarette smoking status and diet (vegetarian/non-vegetarian). No information about physical health was provided. Lipid profile parameters: TC.



Cai et al. (2012)

BMI = body-mass index, FENP = first-episode non-affective psychosis, FES = first-episode schizophrenia; HDL = high-density lipoproteins, LDL = low-density lipoproteins, MINI = Mini-International Neuropsychiatric Interview, OPCRIT = Operational Criteria for Psychotic Illness Checklist, SCID = Structured Clinical Interview for DSM, TC = total cholesterol, TG = triglycerides, WHR = waist-to-hip ratio. a LDL levels determined in 8 patients.

5

6


Fig. 2. Pooled analysis of total cholesterol levels.

was significantly higher in the group of patients compared to controls. Only in some studies (Chen et al., 2016; Misiak et al., 2016a; Ryan et al., 2003; Saddichha et al., 2008; Sengupta et al., 2008; Wu et al., 2013), waist circumference and/or waist-to-hip ratio (WHR) were measured. In the study by Sengupta et al. (2008), WHR was significantly higher in patients with psychosis compared to controls. However, authors found no significant differences in waist circumference between both groups. Patients were diagnosed in accordance with DSM-IV criteria, except for two studies (Hepgul et al., 2012; Nyboe et al., 2015), which used ICD10 criteria. Only some studies used the Structured Clinical Interview for DSM-IV (Chen et al., 2016; Petrikis et al., 2015; Sengupta et al., 2008;

Spelman et al., 2007; Wu et al., 2013), the Operational Criteria for Psychotic Illness checklist (Misiak et al., 2016a) and the Mini-International Neuropsychiatric Interview (Venkatasubramanian et al., 2007). In the majority of studies, only patients with a diagnosis of schizophrenia were included except for three studies, which recruited patients with other diagnostic categories of non-affective psychosis (Hepgul et al., 2012; Petrikis et al., 2015; Sengupta et al., 2008). Only in two studies, a diagnosis of schizophrenia was confirmed after 3 months (Chen et al., 2016) or 6 months (Wu et al., 2013). Controls were recruited through advertising in the local area (Chen et al., 2016; Hepgul et al., 2012; Kirkpatrick et al., 2010; Nyboe et al., 2015; Ryan et al., 2003; Sengupta et al., 2008) and university employees

Fig. 3. Pooled analysis of LDL levels.



or students (Misiak et al., 2016a; Petrikis et al., 2015; Ryan et al., 2003; Sengupta et al., 2008; Wu et al., 2013), from local enterprise employees (Petrikis et al., 2015; Wu et al., 2013), from the hospital staff and their friends or relatives (Enez Darcin et al., 2015; Misiak et al., 2016a; Ryan et al., 2003; Sengupta et al., 2008; Venkatasubramanian et al., 2007; Wu et al., 2013), from patients' relatives (Kavzoglu and Hariri, 2013) and friends (Dasgupta et al., 2010; Kavzoglu and Hariri, 2013), from the register of healthy volunteers (Hepgul et al., 2012) as well as by the word of mouth (Misiak et al., 2016a). In one study, FENP patients and controls were recruited from the mandatory military service (Basoglu et al., 2010). In other studies, information about the way controls were enrolled was not provided (McCreadie and the Scottish Schizophrenia Research Group, 2000; Cai et al., 2012; Saddichha et al., 2008; Spelman et al., 2007; Zhang et al., 2016). Overall, controls had no physical health impairments. In three studies (McCreadie and the Scottish Schizophrenia Research Group, 2000; Hepgul et al., 2012; Saddichha et al., 2008), information about physical health of controls was not provided. Total cholesterol level was measured in 17 studies (Table 1). Heterogeneity was not significant for all studies analysing total cholesterol levels (Q = 19.19, df = 16, p = 0.259, I2 = 16.65%). The pooled data analysis under fixed-effects model revealed significantly lower total cholesterol levels in patients with FENP compared to controls [ES = −0.16 (95% CI: −0.27, −0.06), p = 0.003] (Fig. 2). In subgroup analysis (see Appendix), the overall ES remained significant [− 0.17 (95% CI: −0.30, −0.04), p = 0.012] after removing all studies with significant differences in at least one of potential confounding variables (age, sex, BMI and cigarette smoking) (McCreadie and the Scottish Schizophrenia Research Group, 2000; Enez Darcin et al., 2015; Nyboe et al., 2015; Sengupta et al., 2008; Zhang et al., 2016). Similarly, the ES estimate remained significant after removing each single study. Results of tests for asymmetry were non-significant [the Begg & Mazumbar test: tau = −0.12, Z = −0.66, p = 0.510; the Egger's test: regression intercept = 0.004 (95% CI: −2.45, 2.46), p = 0.997] (see Appendix for funnel plot of asymmetry). Meta-regression analysis revealed that neither the NOS score (B = 0.01, SE = 0.04, Z = 0.29, p = 0.774) nor differences in potential confounding variables (B b 0.01, SE = 0.12, Z b 0.01, p = 0.997) were associated with the magnitude of ES.

7

The levels of LDL were determined in 15 studies (Table 1). Heterogeneity was not significant for all studies measuring LDL levels (Q = 11.81, df = 14, p = 0.621, I2 b 0.01%). The pooled data analysis under fixed-effects model demonstrated significantly lower LDL levels in patients with FENP compared to controls [ES = − 0.13 (95% CI: − 0.24, 0.01), p = 0.034] (Fig. 3). Sensitivity analysis revealed that after removal of single studies (Dasgupta et al., 2010; Kirkpatrick et al., 2010; Misiak et al., 2016a; Ryan et al., 2003; Wu et al., 2013; Zhang et al., 2016), the pooled ES estimate was not significant. However, the ES estimate remained significant [ES = −0.20 (95% CI: −0.34, −0.05), p = 0.007] after removing studies with significant differences between FENP patients and controls in at least one of potential confounding variables (see Appendix) (Enez Darcin et al., 2015; Nyboe et al., 2015; Sengupta et al., 2008; Spelman et al., 2007). Results of tests for asymmetry were nonsignificant [the Begg & Mazumbar test: tau = 0.12, Z = 0.64, p = 0.520; the Egger's test: regression intercept = − 0.222 (95% CI: −1.73, 2.17), p = 0.810] (see Appendix for funnel plot of asymmetry). Meta-regression analysis revealed that neither the NOS score (B b 0.01, SE = 0.05, Z = 0.07, p = 0.943) nor differences in potential confounding variables (B = 0.18, SE = 0.14, Z = 1.30, p = 0.195) were associated with the magnitude of ES. The levels of HDL were measured in 17 studies (Table 1). Heterogeneity was significant (Q = 62.45, df = 16, p b 0.001, I2 = 74.38%) and was led by two outliers according to sensitivity analysis (Basoglu et al., 2010; Wu et al., 2013). Before removal of these studies, results of pooled analysis were significant under random-effects model [ES = − 0.27 (95% CI: −0.49, −0.05), p = 0.018] (Fig. 4). After removal of these studies, heterogeneity was improved (Q = 19.65, df = 14, p = 0.141, I2 = 28.76%) and results of pooled analysis remained significant under fixed-effects model [ES = − 0.29 (95% CI: − 0.41, − 0.18), p b 0.001]. Similarly, the overall ES estimate remained significant [ES = − 0.20 (95% CI: −0.34, −0.06), p = 0.004] after removing all studies with significant differences in at least one of potential confounding variables (age, sex, BMI and cigarette smoking) (Enez Darcin et al., 2015; Nyboe et al., 2015; Sengupta et al., 2008; Spelman et al., 2007; Zhang et al., 2016) and two outliers (Basoglu et al., 2010; Wu et al., 2013) (see Appendix). Results of pooled analysis were also significant after removing any single study. Results of tests for asymmetry were non-significant

Fig. 4. Pooled analysis of HDL levels.


8


Fig. 5. Pooled analysis of triglycerides levels.

[the Begg & Mazumbar test: tau b 0.01, Z b 0.01, p = 1.000; the Egger's test: regression intercept = 1.318 (95% CI: − 2.81, 5.45), p = 0.506] (see Appendix for funnel plot of asymmetry). Meta-regression analysis revealed that neither the NOS score (B = − 0.07, SE = 0.09, Z = −0.80, p = 0.423) nor differences in potential confounding variables (B = −0.33, SE = 0.26, Z = −1.29, p = 0.197) were associated with the magnitude of ES. The levels of triglycerides were determined in 18 studies (Table 1). There was no significant heterogeneity across studies measuring the levels of triglycerides (Q = 11.37, df = 17, p = 0.837, I2 b 0.01%). Pooled analysis under fixed-effects model revealed significantly higher levels of triglycerides in FENP patients compared to controls [ES = 0.22 (95% CI: 0.11, 0.32), p b 0.001] (Fig. 5). In sensitivity analysis, the overall ES estimate remained significant after removing each single study. Similarly, the ES estimate remained significant [ES = 0.22 (95% CI: 0.10–0.34), p b 0.001] after removing studies with significant differences between FENP patients and controls in at least one of potential confounding variables (age, sex, BMI and cigarette smoking) (Enez Darcin et al., 2015; Nyboe et al., 2015; Sengupta et al., 2008; Spelman et al., 2007; Zhang et al., 2016) (see Appendix). Results of tests for asymmetry were nonsignificant [the Begg & Mazumbar test: tau = − 0.12, Z = − 0.72, p = 0.471; the Egger's test: regression intercept = − 0.260 (95% CI: −2.01, 1.49), p = 0.756] (see Appendix for funnel plot of asymmetry). Neither the NOS score (B = −0.05, SE = 0.04, Z = −1.35, p = 0.178) nor differences in potential confounding variables (B = −0.004, SE = 0.12, Z = −0.03, p = 0.974) were associated with the magnitude of ES. 4. Discussion In this meta-analysis, we found significantly lower levels of total cholesterol, LDL and HDL as well as significantly higher levels of triglycerides in antipsychotic-naïve patients with FENP compared to controls, with relatively low ES estimates ranging from 0.13 to 0.27. However, the association between low LDL levels and FENP appeared to be non-significant in sensitivity analysis. Importantly, our results were not associated with significant differences in potential confounding variables, such as age, BMI or cigarette smoking that were present in single studies included in this meta-analysis (McCreadie and the Scottish Schizophrenia Research Group, 2000; Enez Darcin et al., 2015; Nyboe et al., 2015;

Sengupta et al., 2008; Spelman et al., 2007; Zhang et al., 2016). These findings indicate that patients in the early phase of schizophrenia-spectrum disorders present specific lipid disturbances with both anti-atherogenic (low levels of total cholesterol and LDL) and pro-atherogenic (low levels of HDL and high levels of triglycerides) characteristics that most likely fall beyond the impact of environmental factors. These alterations have been recently addressed as potential confounding factors in a meta-analysis addressing abnormal glucose homeostasis in patients with first-episode psychosis (Perry et al., 2016). Authors found significantly lower levels of total cholesterol and HDL. However, it should be noted that our meta-analysis, as opposed to the meta-analysis by Perry et al. (2016), was limited to patients with non-affective psychosis. Therefore, alterations in lipid profile parameters demonstrated in our meta-analysis suggest that schizophrenia-spectrum disorders and dyslipidaemia might share overlapping genetic backgrounds. In support of our results, a recent analysis of data from GWASs revealed that the levels of LDL and HDL together with schizophrenia risk share common genetic risk factors (Andreassen et al., 2013). Furthermore, it has been reported that the methylenetetrahydrofolate reductase (MTHFR) gene polymorphism (C677T) might confer schizophrenia risk (Yadav et al., 2016), predict antipsychotic-induced weight gain (Kao et al., 2014; Misiak et al., 2016b) and increase cardiovascular risk in the general population (Kim and Becker, 2003). Our findings also raise the discussion whether peripheral blood lipid profile disturbances are related to a broader spectrum of lipid abnormalities reported in schizophrenia. Indeed, there are studies showing that patients with schizophrenia present various abnormalities in lipid homeostasis that involve fatty acids and phospholipids, sphingolipids as well as endocannabinoid lipids (Castillo et al., 2016; Schneider et al., 2016). There is also evidence for myelin abnormalities in psychotic disorders that can be attributed to genetic factors (Mighdoll et al., 2015). These alterations have been also observed in magnetic resonance spectroscopy studies of antipsychotic-naïve patients with first-episode psychosis (Miller et al., 2012; Smesny et al., 2015). Importantly, lipid pattern abnormalities in membranes are observed not only within the central nervous system, but also in peripheral cells, such as red blood cells (Tessier et al., 2016) and fibroblasts (Mahadik et al., 1994). For instance, Mahadik et al. (1994) demonstrated that drug-naïve patients at the onset of psychosis had significantly lower levels of total cholesterol,



cholesteryl esters and phospholipids in plasma membranes of skin fibroblasts. Membrane lipid abnormalities have been also associated with more robust cognitive impairments in patients with schizophrenia (Solberg et al., 2015; Tessier et al., 2016). Mechanisms linking central and peripheral lipid abnormalities in schizophrenia remain unknown. Several lines of evidence indicate that there is a cross-talk between central and peripheral lipid metabolism. About 25% of the total cholesterol amount is located within the brain, mainly in the myelin sheaths (Bjorkhem and Meaney, 2004). The rest of brain cholesterol builds up cellular membranes of astrocytes and neurons maintaining cellular integrity and synaptic transmission (Dietschy and Turley, 2004; Zhang and Liu, 2015). However, alterations in the brain cholesterol synthesis and turnover in patients with schizophrenia have not been addressed so far. There is only one study showing lower levels of cholesteryl esters in prefrontal cortex of patients with schizophrenia (Taha et al., 2013). Importantly, cholesterol cannot pass through the blood-brain barrier and thus it is mostly synthesized de novo in the brain. However, excessive cholesterol synthesis is regulated by complex efflux mechanisms, which can link central and peripheral cholesterol concentrations (Bjorkhem, 2002; Vitali et al., 2014; Zhang and Liu, 2015). Another possibility is that our findings might also reflect premorbid exposure to environmental factors. There are studies showing poor dietary habits that manifest in low intake of micronutrients and proteins together with higher consumption of calories and saturated fatty acids (Manzanares et al., 2014; Williamson et al., 2015). In addition, accumulating evidence indicates that polyunsaturated fatty acids (PUFAs) might be also involved in the etiology of psychotic disorders. Although there is a general consensus that PUFAs might be beneficial in terms of improving lipid profile alterations (Jump et al., 2012), studies investigating their role in psychosis have provided mixed results. Indeed, there are studies showing higher dietary intake of omega-3 and omega-6 PUFAs in subjects at risk of psychosis or individuals with psychoticlike experiences (Hedelin et al., 2010; Pawelczyk et al., 2016; Pawelczyk et al., 2015). On the other hand, it has been reported that supplementation of PUFAs might improve clinical outcomes of at-risk individuals or patients with psychosis (Chen et al., 2015; van der Gaag et al., 2013). Other unhealthy lifestyle characteristics, including sedentary behaviour and cigarette smoking, have been also demonstrated in the premorbid period of psychosis (Carney et al., 2016; Mittal et al., 2013). Regardless of whether observed lipid profile alterations occur due to inherent and disease-specific or environmental mechanisms, it is important to note that an intact lipid metabolism might be important for brain development and maturational processes that capture critical aspects schizophrenia pathophysiology (Berger et al., 2006). Our meta-analysis has some limitations that require further discussion. Firstly, it should be noted that studies included in our meta-analysis did not control for a number of factors that might also impact blood lipid profile, such as dietary intake or exercise activity. In addition, only a few studies included the measures of visceral fatty tissue, such as waist circumference or WHR, which are also related to lipid profile alterations. Another limitation is that socioeconomic status, which might be a proxy measure environmental factors, was not determined in studies included in our meta-analysis. In addition, community controls were recruited only in some studies. Therefore, caution should be taken in interpreting the extent of intrinsic links between lipid profile abnormalities and schizophrenia-spectrum disorders. It is also important to note that none of included studies reported an initial number of individuals, who had been approached to participate and reasons of non-inclusion were not provided. Thus, conclusions on sample representativeness cannot be made. Finally, there were various methods of case identification and validation of diagnosis. It should be noted that a diagnostic construct of FENP entails some simplistic conceptualization since only some cases represent true cases of schizophrenia. This point should be also addressed in future studies investigating biological underpinnings of metabolic dysregulation in early psychosis.

9

In conclusion, our results suggest that schizophrenia-spectrum disorders and lipid profile disturbances may share overlapping genetic characteristics or appear as the consequence of premorbid environmental exposure. Future studies should investigate the contribution of these factors and disentangle whether peripheral blood lipid profile alterations are related to central indices of metabolic dysregulation observed in patients with schizophrenia-spectrum disorders. Our findings also imply that accurate monitoring of cardio-metabolic risk should be implemented in the early management of patients with psychosis.

Role of the funding source None.

Contributors BM – registration of meta-analysis protocol, online search, extraction of data, statistical analysis, manuscript writing, BS – online search, ŁŁ – consultation of statistical analysis, DF – consultation of search strategy, manuscript writing. Conflict of interest We declare no competing interests.

Acknowledgement None.

Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.schres.2017.03.031. References Andreassen, O.A., Djurovic, S., Thompson, W.K., Schork, A.J., Kendler, K.S., O'Donovan, M.C., Rujescu, D., Werge, T., van de Bunt, M., Morris, A.P., McCarthy, M.I., Roddey, J.C., McEvoy, L.K., Desikan, R.S., Dale, A.M., 2013. Improved detection of common variants associated with schizophrenia by leveraging pleiotropy with cardiovascular-disease risk factors. Am. J. Hum. Genet. 92 (2), 197–209. Baller, J.B., McGinty, E.E., Azrin, S.T., Juliano-Bult, D., Daumit, G.L., 2015. Screening for cardiovascular risk factors in adults with serious mental illness: a review of the evidence. BMC Psychiatry 15, 55. Basoglu, C., Oner, O., Ates, A.M., Algul, A., Semiz, U.B., Ebrinc, S., Cetin, M., Ozcan, O., Ipcioglu, O.M., 2010. Association between symptom improvement and change of body mass index, lipid profile, and leptin, ghrelin, and cholecystokinin levels during 6-week olanzapine treatment in patients with first-episode psychosis. J. Clin. Psychopharmacol. 30 (5), 636–638. Berger, G.E., Smesny, S., Amminger, G.P., 2006. Bioactive lipids in schizophrenia. Int. Rev. Psychiatry 18 (2), 85–98. Bjorkhem, I., 2002. Do oxysterols control cholesterol homeostasis? J. Clin. Investig. 110 (6), 725–730. Bjorkhem, I., Meaney, S., 2004. Brain cholesterol: long secret life behind a barrier. Arterioscler. Thromb. Vasc. Biol. 24 (5), 806–815. Cai, H.L., Li, H.D., Yan, X.Z., Sun, B., Zhang, Q., Yan, M., Zhang, W.Y., Jiang, P., Zhu, R.H., Liu, Y.P., Fang, P.F., Xu, P., Yuan, H.Y., Zhang, X.H., Hu, L., Yang, W., Ye, H.S., 2012. Metabolomic analysis of biochemical changes in the plasma and urine of first-episode neuroleptic-naive schizophrenia patients after treatment with risperidone. J. Proteome Res. 11 (8), 4338–4350. Carney, R., Cotter, J., Bradshaw, T., Firth, J., Yung, A.R., 2016. Cardiometabolic risk factors in young people at ultra-high risk for psychosis: a systematic review and meta-analysis. Schizophr. Res. 170 (2–3), 290–300. Castillo, R.I., Rojo, L.E., Henriquez-Henriquez, M., Silva, H., Maturana, A., Villar, M.J., Fuentes, M., Gaspar, P.A., 2016. From molecules to the clinic: linking schizophrenia and metabolic syndrome through sphingolipids metabolism. Front. Neurosci. 10, 488. Chen, A.T., Chibnall, J.T., Nasrallah, H.A., 2015. A meta-analysis of placebo-controlled trials of omega-3 fatty acid augmentation in schizophrenia: possible stage-specific effects. Ann. Clin. Psychiatry 27 (4), 289–296. Chen, D.C., Du, X.D., Yin, G.Z., Yang, K.B., Nie, Y., Wang, N., Li, Y.L., Xiu, M.H., He, S.C., Yang, F.D., Cho, R.Y., Kosten, T.R., Soares, J.C., Zhao, J.P., Zhang, X.Y., 2016. Impaired glucose tolerance in first-episode drug-naive patients with schizophrenia: relationships with clinical phenotypes and cognitive deficits. Psychol. Med. 46 (15), 3219–3230. Cordes, J., Bechdolf, A., Engelke, C., Kahl, K.G., Balijepalli, C., Losch, C., Klosterkotter, J., Wagner, M., Maier, W., Heinz, A., de Millas, W., Gaebel, W., Winterer, G., Janssen, B., Schmidt-Kraepelin, C., Schneider, F., Lambert, M., Juckel, G., Wobrock, T., Riedel, M., Moebus, S., 2016. Prevalence of metabolic syndrome in female and male patients at risk of psychosis. Schizophr. Res. Dasgupta, A., Singh, O.P., Rout, J.K., Saha, T., Mandal, S., 2010. Insulin resistance and metabolic profile in antipsychotic naive schizophrenia patients. Prog. NeuroPsychopharmacol. Biol. Psychiatry 34 (7), 1202–1207.


10


Dietschy, J.M., Turley, S.D., 2004. Thematic review series: brain lipids. Cholesterol metabolism in the central nervous system during early development and in the mature animal. J. Lipid Res. 45 (8), 1375–1397. Enez Darcin, A., Yalcin Cavus, S., Dilbaz, N., Kaya, H., Dogan, E., 2015. Metabolic syndrome in drug-naive and drug-free patients with schizophrenia and in their siblings. Schizophr. Res. 166 (1–3), 201–206. Fernandes, B.S., Steiner, J., Bernstein, H.G., Dodd, S., Pasco, J.A., Dean, O.M., Nardin, P., Goncalves, C.A., Berk, M., 2016. C-reactive protein is increased in schizophrenia but is not altered by antipsychotics: meta-analysis and implications. Mol. Psychiatry 21 (4), 554–564. van der Gaag, M., Smit, F., Bechdolf, A., French, P., Linszen, D.H., Yung, A.R., McGorry, P., Cuijpers, P., 2013. Preventing a first episode of psychosis: meta-analysis of randomized controlled prevention trials of 12 month and longer-term follow-ups. Schizophr. Res. 149 (1–3), 56–62. Greenhalgh, A.M., Gonzalez-Blanco, L., Garcia-Rizo, C., Fernandez-Egea, E., Miller, B., Arroyo, M.B., Kirkpatrick, B., 2016. Meta-analysis of glucose tolerance, insulin, and insulin resistance in antipsychotic-naive patients with nonaffective psychosis. Schizophr. Res. Hedelin, M., Lof, M., Olsson, M., Lewander, T., Nilsson, B., Hultman, C.M., Weiderpass, E., 2010. Dietary intake of fish, omega-3, omega-6 polyunsaturated fatty acids and vitamin D and the prevalence of psychotic-like symptoms in a cohort of 33,000 women from the general population. BMC Psychiatry 10, 38. Hepgul, N., Pariante, C.M., Dipasquale, S., DiForti, M., Taylor, H., Marques, T.R., Morgan, C., Dazzan, P., Murray, R.M., Mondelli, V., 2012. Childhood maltreatment is associated with increased body mass index and increased C-reactive protein levels in first-episode psychosis patients. Psychol. Med. 42 (9), 1893–1901. Horrobin, D.F., Manku, M.S., Hillman, H., Iain, A., Glen, M., 1991. Fatty acid levels in the brains of schizophrenics and normal controls. Biol. Psychiatry 30 (8), 795–805. Jump, D.B., Depner, C.M., Tripathy, S., 2012. Omega-3 fatty acid supplementation and cardiovascular disease. J. Lipid Res. 53 (12), 2525–2545. Kao, A.C., Rojnic Kuzman, M., Tiwari, A.K., Zivkovic, M.V., Chowdhury, N.I., Medved, V., Kekin, I., Zai, C.C., Lieberman, J.A., Meltzer, H.Y., Bozina, T., Bozina, N., Kennedy, J.L., Sertic, J., Muller, D.J., 2014. Methylenetetrahydrofolate reductase gene variants and antipsychotic-induced weight gain and metabolic disturbances. J. Psychiatr. Res. 54, 36–42. Kavzoglu, S.O., Hariri, A.G., 2013. Intracellular adhesion molecule (ICAM-1), vascular cell adhesion molecule (VCAM-1) and E-selectin levels in first episode schizophrenic patients. Bull. Clin. Psychopharmacol. 23 (3), 205–214. Kim, R.J., Becker, R.C., 2003. Association between factor V Leiden, prothrombin G20210A, and methylenetetrahydrofolate reductase C677T mutations and events of the arterial circulatory system: a meta-analysis of published studies. Am. Heart J. 146 (6), 948–957. Kirkpatrick, B., Garcia-Rizo, C., Tang, K., Fernandez-Egea, E., Bernardo, M., 2010. Cholesterol and triglycerides in antipsychotic-naive patients with nonaffective psychosis. Psychiatry Res. 178 (3), 559–561. Laursen, T.M., Nordentoft, M., Mortensen, P.B., 2014. Excess early mortality in schizophrenia. Annu. Rev. Clin. Psychol. 10, 425–448. Lawrence, D., Kisely, S., 2010. Inequalities in healthcare provision for people with severe mental illness. J. Psychopharmacol. 24 (4 Suppl), 61–68. de Leon, J., Diaz, F.J., 2005. A meta-analysis of worldwide studies demonstrates an association between schizophrenia and tobacco smoking behaviors. Schizophr. Res. 76 (2– 3), 135–157. Mahadik, S.P., Mukherjee, S., Correnti, E.E., Kelkar, H.S., Wakade, C.G., Costa, R.M., Scheffer, R., 1994. Plasma membrane phospholipid and cholesterol distribution of skin fibroblasts from drug-naive patients at the onset of psychosis. Schizophr. Res. 13 (3), 239–247. Manzanares, N., Monseny, R., Ortega, L., Montalvo, I., Franch, J., Gutierrez-Zotes, A., Reynolds, R.M., Walker, B.R., Vilella, E., Labad, J., 2014. Unhealthy lifestyle in early psychoses: the role of life stress and the hypothalamic-pituitary-adrenal axis. Psychoneuroendocrinology 39, 1–10. McCreadie and the Scottish Schizophrenia Research Group, 2000. Smoking habits and plasma lipid peroxide and vitamin E levels in never-treated first-episode patients with schizophrenia. Br. J. Psychiatry 176, 290–293. McNamara, R.K., Jandacek, R., Rider, T., Tso, P., Hahn, C.G., Richtand, N.M., Stanford, K.E., 2007. Abnormalities in the fatty acid composition of the postmortem orbitofrontal cortex of schizophrenic patients: gender differences and partial normalization with antipsychotic medications. Schizophr. Res. 91 (1–3), 37–50. Mighdoll, M.I., Tao, R., Kleinman, J.E., Hyde, T.M., 2015. Myelin, myelin-related disorders, and psychosis. Schizophr. Res. 161 (1), 85–93. Miller, J., Drost, D.J., Jensen, E., Manchanda, R., Northcott, S., Neufeld, R.W., Menon, R., Rajakumar, N., Pavlosky, W., Densmore, M., Schaefer, B., Williamson, P., 2012. Progressive membrane phospholipid changes in first episode schizophrenia with high field magnetic resonance spectroscopy. Psychiatry Res. 201 (1), 25–33. Misiak, B., Laczmanski, L., Sloka, N.K., Szmida, E., Piotrowski, P., Loska, O., Slezak, R., Kiejna, A., Frydecka, D., 2016a. Metabolic dysregulation in first-episode schizophrenia patients with respect to genetic variation in one-carbon metabolism. Psychiatry Res. 238, 60–67. Misiak, B., Laczmanski, L., Sloka, N.K., Szmida, E., Slezak, R., Piotrowski, P., Kiejna, A., Frydecka, D., 2016b. Genetic variation in one-carbon metabolism and changes in metabolic parameters in first-episode schizophrenia patients. Int. J. Neuropsychopharmacol. Mitchell, A.J., Vancampfort, D., Sweers, K., van Winkel, R., Yu, W., De Hert, M., 2013. Prevalence of metabolic syndrome and metabolic abnormalities in schizophrenia and related disorders—a systematic review and meta-analysis. Schizophr. Bull. 39 (2), 306–318. Mittal, V.A., Gupta, T., Orr, J.M., Pelletier-Baldelli, A., Dean, D.J., Lunsford-Avery, J.R., Smith, A.K., Robustelli, B.L., Leopold, D.R., Millman, Z.B., 2013. Physical activity level and

medial temporal health in youth at ultra high-risk for psychosis. J. Abnorm. Psychol. 122 (4), 1101–1110. Moher, D., Liberati, A., Tetzlaff, J., Altman, D.G., 2009. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. BMJ 339, b2535. Muntjewerff, J.W., Kahn, R.S., Blom, H.J., den Heijer, M., 2006. Homocysteine, methylenetetrahydrofolate reductase and risk of schizophrenia: a meta-analysis. Mol. Psychiatry 11 (2), 143–149. Nyboe, L., Vestergaard, C.H., Moeller, M.K., Lund, H., Videbech, P., 2015. Metabolic syndrome and aerobic fitness in patients with first-episode schizophrenia, including a 1-year follow-up. Schizophr. Res. 168 (1–2), 381–387. Pawelczyk, T., Trafalska, E., Pawelczyk, A., Kotlicka-Antczak, M., 2015. Differences in omega-3 and omega-6 polyunsaturated fatty acid consumption in people at ultrahigh risk of psychosis, first-episode schizophrenia, and in healthy controls. Early Interv. Psychiatry. Pawelczyk, T., Trafalska, E., Kotlicka-Antczak, M., Pawelczyk, A., 2016. The association between polyunsaturated fatty acid consumption and the transition to psychosis in ultra-high risk individuals. Prostaglandins Leukot. Essent. Fat. Acids 108, 30–37. Perry, B.I., McIntosh, G., Weich, S., Singh, S., Rees, K., 2016. The association between firstepisode psychosis and abnormal glycaemic control: systematic review and metaanalysis. Lancet Psychiatry 3 (11), 1049–1058. Petrikis, P., Tigas, S., Tzallas, A.T., Papadopoulos, I., Skapinakis, P., Mavreas, V., 2015. Parameters of glucose and lipid metabolism at the fasted state in drug-naive first-episode patients with psychosis: evidence for insulin resistance. Psychiatry Res. 229 (3), 901–904. Ryan, M.C., Collins, P., Thakore, J.H., 2003. Impaired fasting glucose tolerance in first-episode, drug-naive patients with schizophrenia. Am. J. Psychiatry 160 (2), 284–289. Saddichha, S., Manjunatha, N., Ameen, S., Akhtar, S., 2008. Metabolic syndrome in first episode schizophrenia - a randomized double-blind controlled, short-term prospective study. Schizophr. Res. 101 (1–3), 266–272. Schneider, M., Levant, B., Reichel, M., Gulbins, E., Kornhuber, J., Muller, C.P., 2016. Lipids in psychiatric disorders and preventive medicine. Neurosci. Biobehav. Rev. Sengupta, S., Parrilla-Escobar, M.A., Klink, R., Fathalli, F., Ying Kin, N., Stip, E., Baptista, T., Malla, A., Joober, R., 2008. Are metabolic indices different between drug-naive firstepisode psychosis patients and healthy controls? Schizophr. Res. 102 (1–3), 329–336. Smesny, S., Gussew, A., Biesel, N.J., Schack, S., Walther, M., Rzanny, R., Milleit, B., Gaser, C., Sobanski, T., Schultz, C.C., Amminger, P., Hipler, U.C., Sauer, H., Reichenbach, J.R., 2015. Glutamatergic dysfunction linked to energy and membrane lipid metabolism in frontal and anterior cingulate cortices of never treated first-episode schizophrenia patients. Schizophr. Res. 168 (1–2), 322–329. Solberg, D.K., Bentsen, H., Refsum, H., Andreassen, O.A., 2015. Association between serum lipids and membrane fatty acids and clinical characteristics in patients with schizophrenia. Acta Psychiatr. Scand. 132 (4), 293–300. Spelman, L.M., Walsh, P.I., Sharifi, N., Collins, P., Thakore, J.H., 2007. Impaired glucose tolerance in first-episode drug-naive patients with schizophrenia. Diabet. Med. 24 (5), 481–485. Steen, V.M., Skrede, S., Polushina, T., Lopez, M., Andreassen, O.A., Ferno, J., Hellard, S.L., 2016. Genetic evidence for a role of the SREBP transcription system and lipid biosynthesis in schizophrenia and antipsychotic treatment. Eur. Neuropsychopharmacol. Taha, A.Y., Cheon, Y., Ma, K., Rapoport, S.I., Rao, J.S., 2013. Altered fatty acid concentrations in prefrontal cortex of schizophrenic patients. J. Psychiatr. Res. 47 (5), 636–643. Tessier, C., Sweers, K., Frajerman, A., Bergaoui, H., Ferreri, F., Delva, C., Lapidus, N., Lamaziere, A., Roiser, J.P., De Hert, M., Nuss, P., 2016. Membrane lipidomics in schizophrenia patients: a correlational study with clinical and cognitive manifestations. Transl. Psychiatry 6 (10), e906. Upthegrove, R., Manzanares-Teson, N., Barnes, N.M., 2014. Cytokine function in medication-naive first episode psychosis: a systematic review and meta-analysis. Schizophr. Res. 155 (1–3), 101–108. Vancampfort, D., Stubbs, B., Mitchell, A.J., De Hert, M., Wampers, M., Ward, P.B., Rosenbaum, S., Correll, C.U., 2015. Risk of metabolic syndrome and its components in people with schizophrenia and related psychotic disorders, bipolar disorder and major depressive disorder: a systematic review and meta-analysis. World Psychiatry 14 (3), 339–347. Venkatasubramanian, G., Chittiprol, S., Neelakantachar, N., Naveen, M.N., Thirthall, J., Gangadhar, B.N., Shetty, K.T., 2007. Insulin and insulin-like growth factor-1 abnormalities in antipsychotic-naive schizophrenia. Am. J. Psychiatry 164 (10), 1557–1560. Vitali, C., Wellington, C.L., Calabresi, L., 2014. HDL and cholesterol handling in the brain. Cardiovasc. Res. 103 (3), 405–413. Wells, G.A., Shea, B., O'Connell, D., Peterson, J., Welch, V., Losos, M., Tugwell, P., 2000. The Newcastle-Ottawa Scale (NOS) for Assessing the Quality of Nonrandomised Studies in Meta-Analyses. Williamson, K., Kilner, K., Clibbens, N., 2015. A comparison of the nutrient intake of a community-dwelling first-episode psychosis cohort, aged 19–64 years, with data from the UK population. J. Nutr. Sci. 4, e28. Wu, X., Huang, Z., Wu, R., Zhong, Z., Wei, Q., Wang, H., Diao, F., Wang, J., Zheng, L., Zhao, J., Zhang, J., 2013. The comparison of glycometabolism parameters and lipid profiles between drug-naive, first-episode schizophrenia patients and healthy controls. Schizophr. Res. 150 (1), 157–162. Yadav, U., Kumar, P., Gupta, S., Rai, V., 2016. Role of MTHFR C677T gene polymorphism in the susceptibility of schizophrenia: an updated meta-analysis. Asian J. Psychiatr. 20, 41–51. Zhang, J., Liu, Q., 2015. Cholesterol metabolism and homeostasis in the brain. Protein Cell 6 (4), 254–264. Zhang, B., Deng, L., Wu, H., Lu, X., Peng, L., Wu, R., Guo, W., Chen, J., Li, L., Zhao, J., 2016. Relationship between long-term use of a typical antipsychotic medication by Chinese schizophrenia patients and the bone turnover markers serum osteocalcin and betaCrossLaps. Schizophr. Res. 176 (2–3), 259–263.
