Sage Journals: Discover world-class research

Abstract

Objective:

Recent evidence suggests that adjuvant anti-inflammatory agents could improve the symptoms of patients with schizophrenia. However, the effects of the adjuvant anti-inflammatory agents on cognitive function, general functioning and side effects have not yet been systematically investigated. The present meta-analysis aimed to explore the effects of anti-inflammatory agents in patients with schizophrenia comprehensively.

Method:

We performed a literature search in online databases, including PubMed, EMBASE and the Cochrane Database of Systematic Reviews. Randomized, placebo-controlled double-blind studies that investigated clinical outcomes including psychopathology, neurocognition, general functioning and extrapyramidal side effects were included. The examined anti-inflammatory agents included aspirin, celecoxib, omega-3 fatty acids, estrogen, selective estrogen receptor modulator, pregnenolone, N-acetylcysteine, minocycline, davunetide and erythropoietin.

Results:

Sixty-two double-blind randomized clinical trials studying 2914 patients with schizophrenia met the inclusion criteria for quantitative analysis. Significant overall effects were found for anti-inflammatory agents for reducing total, positive and negative symptom scores in the Positive and Negative Syndrome Scale. Cognitive improvements were significant with minocycline and pregnenolone augmentation therapy. General functioning was significantly enhanced by overall anti-inflammatory agents. There were no significant differences in side effects compared with placebo. Baseline total Positive and Negative Syndrome Scale score and illness duration were identified as moderating factors in the effects of anti-inflammatory augmentation on psychiatric symptom improvements.

Conclusion:

The comparative evaluation of efficacy and safety supported the use of anti-inflammatory adjuvant therapy over the use of antipsychotics alone. However, future studies could focus on patients with homogeneous clinical profile to figure out more detailed effects of anti-inflammatory therapy.

Keywords

Schizophrenia meta-analysis anti-inflammatory agents randomized controlled trials

Introduction

Schizophrenia is a psychiatric illness characterized by psychotic symptoms and accompanied by impairments in cognitive and social functioning (Janicak et al., 2014). The chronic course of this illness, along with its early age of onset, imposes tremendous direct and indirect costs that accumulate to rank schizophrenia as one of the most burdensome disorders worldwide (Rössler et al., 2005). The use of antipsychotic medication is efficacious for positive symptoms; however, reports of treatment resistance and limited response, especially in the realms of negative and cognitive symptoms, are abundant (Dixon et al., 1995; Erhart et al., 2006; Kucharska-Pietura and Mortimer, 2013; Miyamoto et al., 2012; Wiersma et al., 1998). Consequently, the importance of developing alternative or adjuvant treatment options has been advocated.

With evidence accumulating for the adverse role of an excessive inflammatory response in the pathophysiology of schizophrenia, anti-inflammatory agents comprise one of the promising options for adjunctive therapies in schizophrenia (Hanson and Gottesman, 2005; Mitra et al., 2017; Müller et al., 2013). Inflammation in the central nervous system causes neurons to emit inflammatory mediators (e.g. cytokines) that trigger cell death (Ransohoff et al., 2015), and the process of neurodegeneration resulting from such excessive inflammatory responses can lead to abnormal brain morphology (Najjar and Pearlman, 2015). Recent reports on the increased level of pro-inflammatory markers in first-episode psychosis (FEP) and acutely relapsed inpatients have suggested the influence of immune dysfunction in the pathogenesis and progression of schizophrenia (Miller et al., 2011; Upthegrove et al., 2014), and findings from preclinical studies have further corroborated the important role of the inflammatory response in the pathogenesis of schizophrenia (Monji et al., 2009; Torrey and Peterson, 1973). The association between inflammation and risk of schizophrenia has also been demonstrated by longitudinal studies reporting the effects of maternal exposure to infection during pregnancy (Brown and Derkits, 2009; Khandaker et al., 2013; Miller et al., 2013), as well as by a recent genome-wide association study that demonstrated the link between single nucleotide polymorphisms on chromosome 6p22.1, which includes immunity-related genes, and schizophrenia susceptibility (Ripke et al., 2014).

Increasing interest in the use of anti-inflammatory agents as adjunctive therapies for schizophrenia accompanies the need to assess treatment efficacy comprehensively. However, current studies report inconsistent findings. For instance, celecoxib, one of the most widely used nonsteroidal anti-inflammatory drugs (NSAIDs), selectively inhibits cyclooxygenase-2 (COX-2), which is upregulated by pro-inflammatory mediators and induces oxidative stress in brain tissues (Meyer et al., 2011). Significant improvements in the cognitive factor of the Positive and Negative Syndrome Scale (PANSS) have been demonstrated following 5 weeks of celecoxib augmentation (Müller et al., 2005), but no significant advantage of the 8-week celecoxib add-on therapy over antipsychotic monotherapy has been reported (Rapaport et al., 2005). Furthermore, a 26-week intervention with adjunctive omega-3 fatty acids, which regulate inflammatory processes in terms of eicosanoid metabolism and cytokine production (Seibert et al., 1994; Simopoulos, 2002), remarkably improved general psychopathology (Pawelczyk et al., 2016). In addition, 12 weeks of ethyl-eicosapentaenoic acid (E-EPA) supplementation significantly reduced symptoms in patients on clozapine (Peet and Horrobin, 2002). However, no significant effect on symptoms was also reported for 16-week treatment with daily E-EPA augmentation (Fenton et al., 2001).

Current meta-analytical studies examining the clinical effects of anti-inflammatory strategies in schizophrenia also have limitations. These studies have primarily focused on outcome measures indexed by changes in total PANSS or Brief Psychiatric Rating Scale (BPRS) scores (Müller, 2013; Nitta et al., 2013; Sommer et al., 2012, 2013). The effects of anti-inflammatory strategies on other significantly relevant clinical factors, such as cognition, real-life functioning and side effects, remain largely unexamined. The largest meta-analysis conducted to date examined seven anti-inflammatory agents and reported the therapeutic potential of N-acetylcysteine (NAC), estrogens and aspirin for schizophrenia (Sommer et al., 2013). However, the potential moderator variables of effects were not examined by meta-regression. Furthermore, the amelioration was only assessed as to PANSS total scores, and the specific changes in symptom subscales were not discussed. Recently published trials which evaluating the cognitive benefits of adjuvant anti-inflammatory therapy also remain unexamined. With these current limitations in mind, the present study aimed to provide an updated meta-analysis of the treatment efficacy of anti-inflammatory drugs in schizophrenia. We comprehensively examined the therapeutic profile of 10 anti-inflammatory agents, previously reviewed as effective therapeutic options for schizophrenia (Keller et al., 2013; Müller et al., 2013; Sommer et al., 2013), and assessed the efficacy of these treatment options in terms of psychopathology, cognitive functioning, general functioning and side effects. We also implemented a meta-regression framework to examine the possible effects of moderators for the efficacy of anti-inflammatory agents.

Method

Literature search

A systematic search strategy was employed to identify relevant studies. Two independent researchers conducted a two-step literature search. In the first step, a literature search was conducted in PubMed, EMBASE and the Cochrane Database of Systematic Reviews databases. The search spanned from database inception to August 2017, and the following keywords, including their synonyms and combinations, were used as search terms: ‘schizophrenia’, ‘anti-inflammatory agents’, ‘aspirin’, ‘celecoxib’, ‘Omega 3 fatty acids’, ‘EPA’, ‘docosahexaenoic acid (DHA)’, ‘estrogen’, ‘minocycline’, ‘pregnenolone’, ‘N-acetylcysteine’, ‘davunetide’ and ‘erythropoietin’. In the second step, the reference lists of the retrieved studies and reviews were manually searched for additional relevant publications. Eligible articles were selected according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) Statement (Figure 1).

Figure 1.

Preferred Reporting Items for Systematic Reviews and Meta-Analyses flow diagram.

Selection criteria

Studies included in the present meta-analysis met the following criteria: (1) studies published as an original article; (2) randomized, double-blind, placebo-controlled studies regarding antipsychotic augmentation with celecoxib, aspirin, davunetide, minocycline, estrogens (or raloxifene), omega-3 fatty acids, NAC or erythropoietin; (3) studies including patients with a diagnosis of a schizophrenia spectrum disorders; (4) studies providing sufficient information to calculate effect size (ES) statistics of changes in at least one outcome measure of interest and (5) literature written in English. We also included crossover studies in our analysis to obtain as much information as possible. Studies that used additional treatments, cognitive therapy for instance, along with anti-inflammatory agents were excluded. The eligibility of each randomized trial was assessed according to Cochrane’s tool for assessment of the risk of bias.

Data extraction

The following details were collected from the included studies: study location, year of publication, group characteristics, subject characteristics, experimental design, psychometric data, effect estimates and information required to complete the CONSORT checklist (Schulz et al., 2010). Pre- and post-treatment scores and F or p values for treatment effect were extracted when the mean and standard deviation (SD) value of the change scores were not reported (Lipsey and Wilson, 2001). When more than one study shared a study population, we included the one with a larger sample. The authors of the articles were contacted for unpublished information.

Outcome measures

Our primary outcome measures were the mean change (initial minus end-point ratings) in PANSS total and subscale scores. Our second outcome measures included the mean change in the scores of cognitive scales—e.g. MATRICS Consensus Cognitive Battery (MCCB; Nuechterlein et al., 2008), Brief Assessment of Cognition in Schizophrenia (BACS; Keefe et al., 2008), global assessment of functioning (GAF) and side effects measured on the Abnormal Involuntary Movement Scale (AIMS; Gharabawi et al., 2005) or Extrapyramidal Symptom Rating Scale (ESRS; Munetz and Benjamin, 1988). Results for cognitive functioning were grouped by the domains of attention, processing speed, language, visuospatial and constructional skills, visual learning/memory, verbal learning/memory, executive function or social cognition (Supplemental Table 1).

Statistical analysis

For outcome measures, we calculated the ES of the mean differences between the two treatment groups. Significant differences were determined by Hedges’ g test. A random-effects model was employed to take into account the variation between the retrieved studies. Primary analyses including the ES of all trials were conducted, and we then repeated the meta-analysis per drug. The composite of ES of each drug was also presented as a more accurate indicator of the overall effects of anti-inflammatory agents. When a study of cognitive functioning reported data on multiple cognitive tasks classified in the same domain, a combined ES from those tasks was used. The individual study with standardized residual >1.96 for each analysis was considered as an outlier and was excluded from the meta-analysis. Study heterogeneity was measured using I² statistics. In the case of high heterogeneity, meta-regression or subgroup analysis was performed to identify the possible source of the heterogeneity. Potential moderator variables included treatment duration, treatment dose, age, sex, illness duration, antipsychotic dose (olanzapine equivalent) and PANSS total scores at baseline. The presence of publication bias was assessed by visually inspecting funnel plots. If the number of studies was 10 or larger, Egger’s test was conducted to test for asymmetry. To account for significant publication bias, we used the fixed-model trim-and-fill method to estimate the influence of small study effects. All statistical tests were two-sided and used a significance level of p < 0.05. The analyses were performed with Comprehensive Meta-Analysis V2 (Biostat, Inc., Englewood, NJ, USA) or Stata V14 (StataCorp, College Station, TX, USA).

Results

Our literature search yielded 62 eligible trials from 48 articles (PRISMA flowchart, see Figure 1). This search led to a dataset comprising 2914 patients with schizophrenia in total. All included studies compared the efficacy of using anti-inflammatory agents adjunctively with antipsychotics with that of antipsychotic monotherapy and provided information necessary to calculate changes in PANSS scores (k = 57), cognitive functioning (k = 14), GAF scores (k = 9) or extrapyramidal symptoms (k = 15). All included trials were double-blind randomized controlled trials (RCTs), and two of the included trials adopted a crossover design (Ji et al., 2016; Weickert et al., 2015). Study characteristics and patient demographics are listed in Table 1. The methodological quality of each study was evaluated in accordance with the Cochrane Handbook for Systematic Review of Interventions (Supplemental Table 2). Examined drugs included aspirin, celecoxib, omega-3 fatty acids, estrogens, pregnenolone, minocycline, NAC, erythropoietin and davunetide. Publication bias was detected in two of the analyses and was corrected with the trim-and-fill method.

Table 1.

Summary of study characteristics and demographics.

Study	Diagnosis	Treatment		Control		Domains
Study	Diagnosis	N	Age^a	N	Age^a	APS^b	P	C	F	S
Aspirin
Attari et al. (2017)	SZ	40	33.5	20	31.4	–	2
Laan et al. (2010)	SZ/SAD/SFD	33	31.6	37	30.6	–	1	1
Celecoxib
Akhondzadeh et al. (2007)	SZ	30	33.1	30	34.3	–	1			1
Müller et al. (2002)	SZ	25	35.9	25	35.5	–	1
Müller et al. (2010)	SZ/SFD	25	26.2	25	30.9	–	1
Rapaport et al. (2005)	SZ	18	44.1	17	47.3	10.9	1			1
Davunetide
Javitt et al. (2012)	SZ	41	44.2	22	41.4	–		2
Erythropoietin
Ehrenreich et al. (2007)	SZ	19	39.6	18	43	–	1	1
Estrogen
Akhondzadeh et al. (2003)	SZ	16	32.1	16	33.4	–	1			1
Ghafari et al. (2013)	SZ	15	34.2	15	34.8	–	1
Ko et al. (2006)	SZ/SAD	14	33.4	14	35.9	14.6		1
Kulkarni et al. (2001)	SZ/SAD/SFD	12	32.8	12	34.9	–	2
Kulkarni et al. (2008)	SZ/SAD/SFD	51	33.5	36	33.7	–	1
Kulkarni et al. (2011)	SZ/SAD/SFD	26	32.9	27	31.2	–	1
Kulkarni et al. (2015)	SZ/SAD/SFD	56	34.8	62	35.4	–	2
Minocycline
Chaudhry et al. (2012)	SZ/SAD/SFD	46	–	48	–	–	2		2	2
Ghanizadeh et al. (2014)	SZ	21	31	22	30.2	16.99	1
Khodaie-Ardakani et al. (2014)	SZ	20	41.05	20	38.95	11.13	1			1
Levkovitz et al. (2010)	SZ	36	25.1	18	24.7	–	1	1	1	1
Liu et al. (2014)	SZ	39	27.05	40	27.7	9.53	1	1
N-acetylcysteine
Berk et al. (2008)	SZ	69	37.2	71	36.1	16.9	1		1	1
Farokhnia et al. (2013)	SZ	21	32.2	21	33.4	–	1			1
Omega-3 fatty acids
Bentsen et al. (2013)	SZ/SAD/SFD	49	26.1	48	28.5	–	1
Boskovic et al. (2016)	SZ	18	54.2	16	45.7	6.6	2			2
Emsley et al. (2002)	SZ	20	46.2	20	43.6	25	1
Emsley et al. (2006)	SZ/SAD	39	42.4	38	43.4	–	1
Fenton et al. (2001)	SZ/SAD	37	–	38	–	–	1	1		1
Jamilian et al. (2014)	SZ	30	32	30	31	–	1
Manteghiy et al. (2008)	SZ	42	37.4	43	39	–	1
Pawelczyk et al. (2016)	SZ	36	23.2	35	23.3	6.8	1		1
Peet et al. (2001)	SZ	45	40.1	26	40.5	–	3
Peet and Horrobin (2002)	SZ	58	36.3	43	39	–	6
Qiao et al. (2017)	SZ	28	–	22	–	–	1
Wood et al. (2010)	SZ	9	18.7	8	21.6	–	1		1
Pregnenolone
Kreinin et al. (2014)	SZ/SAD	25	26.9	27	23.9	11.1		1
Marx et al. (2009)	SZ/SAD	9	52.7	9	49.5	15.2	1	1
Marx et al. (2014)	SZ	56	36.8	55	38	–		1
Ritsner et al. (2010)	SZ	20	37.1	11	34.6	14.1	2	1	2	2
Ritsner et al. (2014)	SZ/SAD	25	26.9	27	23.9	11.1	1		1
Selective estrogen receptor modulators
Khodaie-Ardakani et al. (2015)	SZ	21	31.4	21	32.4	–	1			1
Kianimehr et al. (2014)	SZ	25	61.96	25	60.44	–	1
Kulkarni et al. (2010)	SZ/SAD/SFD	22	53.8	13	50.9	–	1
Kulkarni et al. (2016)	SZ/SAD	26	52.9	30	53.1	18.7	1
Ji et al. (2016)	SZ/SAD	10	–	10	–	18.1	1
Usall et al. (2011)	SZ	16	60.1	17	62.7	12.8	1
Usall et al. (2016)	SZ	38	62	32	61.3	–	1
Weickert et al. (2015)	SZ/SAD	39	37.4	38	34	13.6	1	1
Weiser et al. (2017)	SZ/SAD	100	56.6	100	55.8	–	1	1

APS: antipsychotics; P: psychopathology; C: cognition; F: general functioning; S: extrapyramidal side effects; SZ: schizophrenia; SAD: schizoaffective disorder; SFD: schizophreniform disorder.

Mean age.

Olanzapine equivalent dose (mg).

Psychopathology

Overall effects of all anti-inflammatory agents

In total, 57 trials investigated the effects of nine anti-inflammatory agents in terms of psychopathology (Table 2). Davunetide was excluded from the present meta-analysis due to the lack of eligible studies. Our meta-analysis revealed the superior efficacy of anti-inflammatory agents in reducing PANSS total (g = 0.41, 95% confidence interval (CI) = [0.26, 0.56]), positive (g = 0.31, 95% CI = [0.14, 0.48]) and negative (g = 0.38, 95% CI = [0.23, 0.52]) scores. Heterogeneity was moderate to high (I² = 74.2%, 75.5%, 66.0%, each), and subsequent meta-regression analysis revealed that baseline total PANSS scores significantly modulated the effects of anti-inflammatory agents for PANSS positive scores (n = 43, p < 0.01). Publication bias was present for PANSS total and positive symptoms (p < 0.05). For total symptoms, trim-and-fill correction identified three missing studies, but the ES was still significant (g = 0.32, 95% CI = [0.16, 0.49]). The ES for positive symptoms also remained the same after correction.

Table 2.

Treatment efficacy of anti-inflammatory agents.

Studies		N	Dose (g)	Weeks	Positive and Negative Syndrome Scale
					Total symptom			Positive symptom			Negative symptom
					ES (g)	95% CI	p	ES (g)	95% CI	p	ES (g)	95% CI	p
Aspirin
Attari et al. (2017)		40	0.325	6	1.57	[0.93, 2.22]	0	1.91	[1.23, 2.6]	0	1.22	[0.6, 1.83]	0
Attari et al. (2017)		40	0.5	6	2.08	[1.32, 2.84]	0	3.04	[2.14, 3.94]	0	1.62	[0.92, 2.33]	0
Laan et al. (2010)		70	1	12	0.37	[–0.1, 0.83]	0.13	0.23	[–0.24, 0.69]	0.34	0.28	[–0.19, 0.74]	0.24
Celecoxib
Akhondzadeh et al. (2007)		60	0.4	8	0.93	[0.41, 1.46]	0	–	–	–	–	–	–
Müller et al. (2002)		50	0.4	5	0.54	[–0.01, 1.1]	0.06	0.37	[–0.18, 0.92]	0.19	0.47	[–0.09, 1.02]	0.1
Müller et al. (2010)		50	0.4	6	0.52	[–0.03, 1.08]	0.06	0.56	[0, 1.11]	0.05	0.4	[–0.15, 0.95]	0.15
Rapaport et al. (2005)		35	0.4	9	−0.34	[–0.99, 0.31]	0.31	−0.38	[–1.04, 0.27]	0.25	−0.19	[–0.84, 0.46]	0.57
Erythropoietin
Ehrenreich et al. (2007)		39	40,000 IU/week	12	–	–	–	−0.48	[–1.12, 0.16]	0.14	−0.14	[–0.75, 0.48]	0.66
Estrogen
Akhondzadeh et al. (2003)		32	0.0005	8	1.7	[0.9, 2.49]	0	0.97	[0.25, 1.68]	0.01	0.89	[0.18, 1.6]	0.01
Ghafari et al. (2013)		30	0.000625	4	3.71	[2.54, 4.88]	0	1.18	[0.42, 1.93]	0	1.51	[0.72, 2.31]	0
Kulkarni et al. (2001)		24	0.00005	4	0.4	[–0.38, 1.18]	0.31	0.08	[–0.69, 0.85]	0.84	0.4	[–0.39, 1.18]	0.32
Kulkarni et al. (2001)		24	0.0001	4	0.62	[–0.18, 1.41]	0.13	0.44	[–0.34, 1.23]	0.27	0.53	[–0.25, 1.32]	0.18
Kulkarni et al. (2008)		87	0.0001	4	0.58	[0.15, 1.02]	0.01	0.61	[0.18, 1.04]	0.01	0.28	[–0.14, 0.7]	0.19
Kulkarni et al. (2011)		53	0.002	2	−0.19	[–0.72, 0.35]	0.5	−0.28	[–0.82, 0.25]	0.3	0.03	[–0.5, 0.56]	0.92
Kulkarni et al. (2015)		118	0.0001	7	0.19	[–0.17, 0.55]	0.31	0.03	[–0.33, 0.39]	0.86	−0.01	[–0.37, 0.35]	0.94
Kulkarni et al. (2015)		124	0.0002	7	0.45	[0.09, 0.8]	0.01	0.09	[–0.26, 0.44]	0.63	0.34	[–0.01, 0.69]	0.06
Minocycline
Chaudhry et al. (2012)^a	A	24	0.2	48	1.74	[0.82, 2.66]	0	1.07	[0.3, 1.92]	0.01	0.38	[0.54, 2.28]	0
Chaudhry et al. (2012)^a	B	70	0.2	48	0.19	[–0.28, 0.65]	0.43	0.22	[–0.24, 0.69]	0.35	1.41	[–0.09, 0.85]	0.11
Ghanizadeh et al. (2014)		43	0.2	8	0.16	[–0.43, 0.75	0.59	−0.35	[–0.95, 0.24]	0.24	0.29	[–0.3, 0.88]	0.34
Khodaie-Ardakani et al. (2014)		40	0.1-0.2	8	1.49	[0.8, 2.18]	0	0.36	[–0.26, 0.97]	0.25	1.78	[1.06, 2.5]	0
Levkovitz et al. (2010)		54	0.2	24	−0.45	[–1.01, 0.11]	0.12	−0.07	[–1.6, –0.42]	0	0.48	[–0.08, 1.05]	0.11
Liu et al. (2014)		79	0.2	16	0.56	[0.11, 1.01]	0.01	0.22	[–0.22, 0.65]	0.33	1.43	[0.94, 1.92]	0
N-acetylcysteine
Berk et al. (2008)		140	2	24	0.45	[0.11, 0.78]	0.01	0.1	[–0.23, 0.43]	0.55	0.42	[0.08, 0.75]	0.01
Farokhnia et al. (2013)		42	1–2	8	1	[0.37, 1.63]	0	−0.23	[–0.83, 0.36]	0.44	1.1	[0.46, 1.74]	0
Omega-3 fatty acids
Bentsen et al. (2013)		97	2	16	−0.36	[–0.75, 0.04]	0.08	−0.42	[–0.82, –0.02]	0.04	–	–	–
Boskovic et al. (2016)^b	A	20	1.1	16	−0.38	[–1.23, 0.47]	0.38	0.12	[–0.73, 0.96]	0.79	−0.22	[–1.06, 0.63]	0.61
Boskovic et al. (2016)^b	B	14	1.1	16	0.04	[–0.98, 1.07]	0.94	0.42	[–0.61, 1.46]	0.42	0.48	[–0.56, 1.52]	0.36
Emsley et al. (2002)		40	3	12	0.53	[–.09, 1.15]	0.09	–	–	–
Emsley et al. (2006)		77	2	12	−0.03	[–0.47, 0.41]	0.9	–	–	–
Fenton et al. (2001)		75	3	16	−0.06	[–0.51, 0.39]	0.8
Jamilian et al. (2014)		60	1	8	0.26	[–0.24, 0.76]	0.31	−0.12	[–0.62, 0.38]	0.65	−0.04	[–0.54, 0.46]	0.89
Pawelczyk et al. (2016)		71	2.2	26	0.58	[0.11, 1.05]	0.02	2.3	[1.7, 2.89]	0	0.25	[–0.21, 0.71]	0.29
Peet et al. (2001)^c	A	30	2	12	−0.12	[–0.82, 0.57]	0.73	0.25	[–0.46, 0.96]	0.5	–	–	–
Peet et al. (2001)^c	B	29	2	12	0.29	[–0.42, 1.01]	0.42	−0.34	[–1.04, 0.37]	0.35	–	–	–
Peet et al. (2001)^c	C	26	2	12	0.28	[–0.47, 1.03]	0.46	0.56	[–0.2, 1.33]	0.15	–	–	–
Peet et al. (2002)^d	A	23	1	12	−0.48	[–1.28, 0.32]	0.24	−0.46	[–1.26, 0.34]	0.26	−0.43	[–1.23, 0.37]	0.29
Peet et al. (2002)^d	B	26	2	12	−0.06	[–0.81, 0.68]	0.87	−0.03	[–0.77, 0.72]	0.95	−0.3	[–1.06, 0.45]	0.43
Peet et al. (2002)^d	C	16	2	12	1.08	[0.07, 2.08]	0.04	–	–	–	–	–	–
Peet et al. (2002)^d	D	21	1	12	−0.53	[–1.38, 0.31]	0.22	−0.55	[–1.39, 0.3]	0.2	–	–	–
Peet et al. (2002)^d	E	17	2	12	−0.3	[–1.3, 0.7]	0.55	−0.22	[–1.22, 0.77]	0.66	−0.35	[–1.35, 0.64]	0.49
Peet et al. (2002)^d	F	22	4	12	−0.19	[–1, 0.62]	0.65	0.12	[–0.68, 0.93]	0.76	−0.07	[–0.88, 0.74]	0.86
Manteghiy et al. (2008)		85	3	6				0.11	[–0.31, 0.53]	0.6	0.22	[–0.2, 0.64]	0.3
Qiao et al. (2017)		50	0.9	4–12	0.04	[–0.51, 0.59]	0.89	–	–	–	–	–	–
Wood et al. (2010)		17	3	6	−0.13	[–1.03, 0.78]	0.79	–	–	–	–	–	–
Pregnenolone
Marx et al. (2009)		18	0.1–0.5	8	0.57	[–0.35, 1.49]	0.23	0.9	[–0.05, 1.86]	0.06	0.51	[–0.41, 1.43]	0.28
Ritsner et al. (2010)		25	0.03	8	–	–	–	0.06	[–0.7, 0.83]	0.87	−0.04	[–0.81, 0.72]	0.91
Ritsner et al. (2010)		17	0.2	8	–	–	–	0.27	[–0.68, 1.22]	0.57	−0.09	[–1.04, 0.85]	0.85
Ritsner et al. (2014)		52	0.05	8	0.51	[–0.04, 1.05]	0.07	−0.04	[–0.57, 0.5]	0.89	0.08	[–0.45, 0.62]	0.77
Selective estrogen receptor modulators
Khodaie-Ardakani et al. (2015)		42	0.12	8	1.51	[0.83, 2.18]	0	0.44	[–0.16, 1.04]	0.15	0.12	[–0.48, 0.71]	0.69
Kianimehr et al. (2014)		50	0.12	8	0.59	[0.03, 1.15]	0.04	0.99	[0.41, 1.57]	0	0.07	[–0.47, 0.62]	0.8
Kulkarni et al. (2010)		22	0.12	12	0.03	[–0.79, 0.85]	0.94	–	–	–	–	–	–
Kulkarni et al. (2016)		56	0.12	12	0.62	[0.09, 1.15]	0.02	0.3	[–0.22, 0.82]	0.26	0.42	[–0.1, 0.95]	0.11
Ji et al. (2016)		20	0.12	6	0.16	[–0.26, 0.59]	0.45	0.43	[–0.01, 0.87]	0.06	0.11	[–0.31, 0.54]	0.6
Usall et al. (2011)		33	0.06	12	0.52	[–0.16, 1.19]	0.13	1.01	[0.3, 1.72]	0.01	0.69	[0.13, 1.25]	0.02
Usall et al. (2016)		70	0.06	24	0.66	[0.18, 1.14]	0.01	0.2	[–0.27, 0.67]	0.4	0.52	[0.04, 0.99]	0.03
Weickert et al. (2015)		77	0.12	6	0.13	[–0.1, 0.36]	0.28	–	–	–	–	–	–
Weiser et al. (2017)		200	0.12	16	−0.18	[–0.46, 0.09]	0.2	−0.09	[–0.36, 0.19]	0.53	−0.21	[–0.48, 0.07]	0.15

ES: effect size; CI: confidence interval.

Notes: ^aStudy site was Brazil in trial A and Pakistan in trial B.

Adjuvant omega-3 with vitamin E was administered in trial B.

DHA was administered in trial A. EPA was administered in trial B and C.

Background medications were atypical antipsychotics in trial A and B, clozapine in trial C and typical antipsychotics in trial D, E and F.

A-F letter represents the subgroup in the study.

When the pooled ESs for each anti-inflammatory agent were combined, the estimated efficacy increased for PANSS total scores (k = 8, g = 0.48, 95% CI = [0.22, 0.73]; see bar plot in Figure 2), while it decreased for positive (k = 9, g = 0.19, 95% CI = [0.03, 0.35]) and negative (k = 9, g = 0.30, 95% CI = [0.10, 0.50]) scores. Heterogeneity was low to moderate (I² = 65.9%, 30.9%, 59.0%, each). The ES was still significant when we narrowed the criteria for the anti-inflammatory agent to celecoxib and aspirin with a direct anti-inflammatory mechanism (g = 0.59, 95% CI = [0.14, 1.04]).

Figure 2.

Effects of anti-inflammatory agents on total PANSS scores.

Aspirin

Three trials evaluated the effects of aspirin in a total of 130 patients (Attari et al., 2017; Laan et al., 2010). Adjunctive use of aspirin was more effective than antipsychotic monotherapy in improving PANSS total (g = 1.31, 95% CI = [0.25, 2.37]), positive (g = 1.69, 95% CI = [0.05, 3.33]) and negative (g = 1.01, 95% CI = [0.18, 1.83]) scores. Heterogeneity was high (I² = 88.8%, 94.5%, 83.0%, each). Subgroup analysis showed that aspirin treatment was more effective for clinical trials with higher baseline PANSS total score, lower treatment dose and shorter treatment duration (Supplemental Table 3).

Celecoxib

Four RCTs investigated the effects of celecoxib in 195 patients (Akhondzadeh et al., 2007; Müller et al., 2002, 2010; Rapaport et al., 2005). There was no significant difference between add-on celecoxib and standard antipsychotic treatments in terms of PANSS total (g = 0.44, 95% CI = [–0.05, 0.93]), positive (g = 0.24, 95% CI = [–0.31, 0.79]) and negative (g = 0.26, 95% CI = [–0.12, 0.64]) scores. Heterogeneity was moderate to low (I² = 66.6%, 62.4%, 11.8%, each).

Omega-3 fatty acids

Twenty trials examined the effects of omega-3 in a total of 778 patients (Boskovic et al., 2016; Emsley et al., 2002, 2006; Fenton et al., 2001; Jamilian et al., 2014; Pawelczyk et al., 2016; Peet et al., 2001; Peet and Horrobin, 2002; Qiao et al., 2017; Wood et al., 2010). Eleven trials administered eicosapentaenoic acid (EPA), one trial administered DHA and six trials administered a mixture of EPA, DHA or oleic acid. Six trials with patients taking different antipsychotics were retrieved from a dose-ranging study (Peet and Horrobin, 2002). No significant treatment efficacy was found for omega-3 in terms of PANSS total (g = 0.03, 95% CI = [–0.14, 0.20]), positive (g = 0.13, 95% CI = [–0.28, 0.55]) and negative (g = 0.03, 95% CI = [–0.18, 0.24]) scores. Heterogeneity was low (I² = 21.4% and 5.4%) for total and negative symptoms but high (I² = 81.2%) for positive symptoms. Meta-regression analysis revealed a trend-level effect of illness duration on changes in positive PANSS scores (n = 6, p = 0.074). Within the range of 0 to 16 years, there was a trend for shorter illness duration yielding larger ESs.

Estrogens

Eight trials assessed the effects of estrogens in 342 female patients with schizophrenia (Akhondzadeh et al., 2003; Ghafari et al., 2013; Kulkarni et al., 2001, 2008, 2011, 2015).

Among studies included for total symptom analysis, one study (Ghafari et al., 2013) with ES of 3.7 was excluded for high standardized residuals (p < 0.05). The results demonstrated the superior efficacy of estrogens in reducing PANSS total (g = 0.47, 95% CI = [0.13, 0.81]), positive (g = 0.33, 95% CI = [0.02, 0.64]) and negative (g = 0.40, 95% CI = [0.11, 0.69]) scores. Heterogeneity was moderate (66.2%, 62.6%, 56.7%, each).

Selective estrogen receptor modulators

Nine trials investigated the effects of selective estrogen receptor modulators (SERMs) in a total of 583 patients (Ji et al., 2016; Khodaie-Ardakani et al., 2015; Kianimehr et al., 2014; Kulkarni et al., 2010, 2016; Usall et al., 2011, 2016; Weickert et al., 2015; Weiser et al., 2017). The included trials all administered raloxifene. Adjuvant raloxifene therapy significantly reduced PANSS total (g = 0.40, 95% CI = [0.10, 0.70]) and positive (g = 0.41, 95% CI = [0.10, 0.72]) but not negative (g = 0.19, 95% CI = [–0.06, 0.45]) scores. While heterogeneity was moderate (I² = 74.0%, 65.6%, 49.8%, each), publication bias was detected from the funnel plot. The trim-and-fill procedure significantly reduced the ESs for the total (g = 0.20, 95% CI = [–0.08, 0.50]), positive (g = 0.12, 95% CI = [–0.11, 0.36]) and negative (g = 0.09, 95% CI = [–0.12, 0.30]) symptom scores.

Pregnenolone

Four trials assessed the effect of add-on pregnenolone therapy for schizophrenia in 101 patients (Marx et al., 2009, 2014; Ritsner et al., 2010). Pregnenolone was more effective than antipsychotic monotherapy in reducing PANSS total (g = 0.52, 95% CI = [0.05, 0.99]) scores, while its effect on positive (g = 0.17, 95% CI = [–0.19, 0.54]) and negative (g = 0.09, 95% CI = [–0.27, 0.46]) symptom scores was not significant. Heterogeneity was low (I² = 0%) for all analyses.

Minocycline

Six trials evaluated the effects of minocycline in a total of 310 patients (Chaudhry et al., 2012; Ghanizadeh et al., 2014; Khodaie-Ardakani et al., 2014; Levkovitz et al., 2010; Liu et al., 2014). The results demonstrated that adjunctive minocycline was more efficacious for reducing PANSS negative (g = 0.93, 95% CI = [0.42, 1.44]) scores than antipsychotic monotherapy. However, there was no significant difference between the two therapies in terms of total (g = 0.56, 95% CI = [–0.01, 1.12]) and positive (g = 0.19, 95% CI = [–0.11, 0.49]) symptom scores. Heterogeneity was high for total and negative symptoms (I² = 82.4%, 42.0%, 77.6%, each). No influential moderator variable was suggested from the subgroup analysis since the heterogeneity remained high in each subgroup (Supplemental Tables 4 and 5).

NAC

Two trials, including 182 patients, assessed the effects of NAC (Berk et al., 2008; Farokhnia et al., 2013). NAC was superior to placebo in PANSS total (g = 0.65, 95% CI = [0.13, 1.18]) and negative (g = 0.70, 95% CI = [–0.04, 1.36]) scores, while the two treatments were not significantly different in terms of positive (g = 0.02, 95% CI = [–0.27, 0.31]) symptoms. Heterogeneity was moderate to low (I² = 56.3%, 0%, 71.03%, respectively).

Erythropoietin

Only one trial using add-on erythropoietin was retrieved; it included 37 patients (Ehrenreich et al., 2007). There were no significant differences between the treatment efficacy of erythropoietin and placebo in terms of PANSS positive (g = –0.48, 95% CI = [–1.12, 0.16]) or negative (g = –0.14, 95% CI = [–0.75, 0.48]) scores.

General functioning (GAF)

GAF scores were reported as an outcome measure in nine trials that administered omega-3 fatty acids, NAC, minocycline or pregnenolone (Berk et al., 2008; Chaudhry et al., 2012; Levkovitz et al., 2010; Pawelczyk et al., 2016; Ritsner et al., 2010, 2014; Wood et al., 2010). Overall, GAF scores were significantly higher in the anti-inflammatory augmentation group than in the control group (g = 0.22, 95% CI = [0.04, 0.40]). However, the ES measured for each agent was not significant. The majority of the included trials reported no significant improvement in functioning except for one 26-week intervention study providing a 2.2-g combination dose of EPA and DHA (g = 0.61, CI = [0.14, 1.08]; Pawelczyk et al., 2016). Heterogeneity was low (I² = 0–20%) for all analyses.

Neurocognition

Fourteen studies on aspirin, omega-3, estrogen, SERM, davunetide, minocycline, pregnenolone or erythropoietin were included in our analysis of cognitive improvement (Table 3). Significant cognitive benefits were found for minocycline (k = 10, g = 0.21, 95% CI = [0.04, 0.38]) and pregnenolone (k = 20, g = 0.19, 95% CI = [0.08, 0.29]).

Table 3.

Outcomes of cognitive tasks in each intervention trial.

Study	Cognitive domains	Effect Size (g)	95% CI	p
Aspirin
Laan et al. (2010)	Attention	−0.33	[–0.84, 0.18]	0.21
	Motor skill	0.09	[–0.27, 0.45]	0.63
	Verbal learning/memory	0.29	[–0.22, 0.8]	0.26
Davunetide
Javitt et al. (2012)	Attention	−0.11	[–0.75, 0.53]	0.74
	Executive function	0.16	[–0.48, 0.8]	0.63
	Social cognition	0.03	[–0.62, 0.67]	0.94
	Processing speed	0.41	[–0.24, 1.06]	0.21
	Verbal learning/memory	−0.86	[–1.53, –0.19]	0.01
	Visual learning/memory	0.67	[0.01, 1.32]	0.05
	Working memory	0.35	[–0.3, 0.99]	0.29
Javitt et al. (2012)	Attention	−0.06	[–0.7, 0.58]	0.86
	Executive function	0.07	[–0.57, 0.71]	0.83
	Social cognition	0.26	[–0.38, 0.9]	0.43
	Processing speed	0.06	[–0.59, 0.7]	0.87
	Verbal learning/memory	−0.29	[–0.93, 0.36]	0.38
	Visual learning/memory	0.18	[–0.46, 0.82]	0.59
	Working memory	0.1	[–0.55, 0.74]	0.77
Erythropoietin
Ehrenreich et al. (2007)	Attention	0.35	[–0.09, 0.79]	0.11
	Executive function	0.05	[–0.57, 0.66]	0.88
	Language	−0.21	[–0.82, 0.41]	0.51
	Motor skill	−0.1	[–0.53, 0.34]	0.67
	Processing speed	0.17	[–0.44, 0.79]	0.58
	Verbal learning/memory	0	[–0.28, 0.28]	1
	Visual learning/memory	0.47	[–0.15, 1.1]	0.14
	Visuospatial/construction	0.13	[–0.31, 0.56]	0.57
Estrogen
Ko et al. (2006)	Executive function	0.48	[–0.25, 1.21]	0.2
	Processing speed	0.46	[0.04, 0.89]	0.03
	Verbal learning/memory	0.45	[–0.28, 1.18]	0.23
	Visual learning/memory	0.13	[–0.59, 0.85]	0.72
Minocycline
Levkovitz et al. (2010)	Attention	0.19	[–0.66, 1.04]	0.66
	Executive function	1.35	[0.41, 2.29]	0.01
	Motor skill	−0.88	[–1.77, 0.01]	0.05
	Visual learning/memory	0.94	[0.3, 1.57]	0.00
Liu et al. (2014)	Attention	0.16	[–0.06, 0.38]	0.16
	Executive function	0.17	[–0.09, 0.42]	0.2
	Motor skill	0.26	[–0.05, 0.57]	0.1
	Processing speed	0.2	[–0.05, 0.45]	0.12
	Verbal learning/memory	0.08	[–0.23, 0.39]	0.63
	Working memory	0.12	[–0.32, 0.55]	0.61
Omega-3 fatty acids
Fenton et al. (2001)	Composite	0.25	[–0.17, 0.66]	0.25
Pregnenolone
Kreinin et al. (2014)	Attention	0.41	[0.01, 0.81]	0.05
	Executive function	0.17	[–0.23, 0.58]	0.41
	Visual learning/memory	0.18	[–0.21, 0.58]	0.36
Marx et al. (2009)	Attention	0.37	[–0.52, 1.26]	0.41
	Executive function	0.13	[–0.5, 0.76]	0.68
	Motor skill	0.46	[–0.44, 1.35]	0.32
	Social cognition	−0.29	[–1.18, 0.59]	0.52
	Processing speed	−0.1	[–0.54, 0.35]	0.67
	Verbal learning/memory	0.42	[–0.21, 1.05]	0.19
	Visual learning/memory	−0.28	[–1.17, 0.6]	0.53
	Working memory	0.14	[–0.38, 0.65]	0.61
Marx et al. (2014)	Attention	−0.06	[–0.43, 0.31]	0.76
	Executive function	0.36	[–0.01, 0.74]	0.06
	Social cognition	0.13	[–0.24, 0.5]	0.48
	Processing speed	0.17	[–0.2, 0.54]	0.37
	Verbal learning/memory	0.12	[–0.25, 0.49]	0.53
	Visual learning/memory	0.14	[–0.23, 0.51]	0.46
	Working memory	0.19	[–0.18, 0.56]	0.31
Ritsner et al. (2010)	Attention	0.77	[0.07, 1.47]	0.03
Ritsner et al. (2010)	Visual learning/memory	0.72	[0.02, 1.41]	0.05
Selective estrogen receptor modulators
Weickert et al. (2015)	Processing speed	0.42	[0.02, 0.82]	0.04
Weiser et al. (2017)	Executive function	0	[–0.28, 0.28]	1
	Motor skill	−0.14	[–0.41, 0.14]	0.33
	Processing speed	0.05	[–0.15, 0.25]	0.62
	Verbal learning/memory	0.08	[–0.2, 0.35]	0.59
	Working memory	−0.12	[–0.4, 0.16]	0.39

CI: confidence interval.

In minocycline augmentation trials, significant improvements were found in the cognitive domains of visual learning (Levkovitz et al., 2010), executive function (Levkovitz et al., 2010) and attention (Liu et al., 2014). For pregnenolone, one large study reported significant improvement in visual learning (Ritsner et al., 2014) and attention (Kreinin et al., 2014; Ritsner et al., 2014). There was no significant cognitive improvement when the effects of omega-3, estrogens, SERM, aspirin, davunetide and erythropoietin were pooled. Heterogeneity was low to moderate (I² = 0–63.4%) for each augmentation strategy.

Side effects (ESRS/AIMS)

ESRS or AIMS scores were reported as outcome measures in 15 trials that administered omega-3 fatty acids (Boskovic et al., 2016; Emsley et al., 2002; Fenton et al., 2001), estrogens (Akhondzadeh et al., 2003; Khodaie-Ardakani et al., 2015), NAC (Berk et al., 2008; Farokhnia et al., 2013), pregnenolone (Ritsner et al., 2010), minocycline (Chaudhry et al., 2012; Khodaie-Ardakani et al., 2014; Levkovitz et al., 2010) or celecoxib (Akhondzadeh et al., 2007; Rapaport et al., 2005). No significant difference in ESRS or AIMS scores between groups (g = 0.07, 95% CI = [–0.08, 0.21]) was revealed by quantitative analysis. Moreover, subgroup analyses for omega-3 fatty acids, estrogens, SERM, NAC, pregnenolone, minocycline and celecoxib did not reveal any significant effects for side effects.

Discussion

We have comprehensively reviewed the efficacy of anti-inflammatory agents with various outcome measures, including psychopathology, neurocognition, general functioning and extrapyramidal side effects. The meta-analytic results of nine drugs showed that adjuvant anti-inflammatory agents could effectively reduce psychiatric symptoms measured on PANSS, with a small to moderate ES. Significant cognitive benefits were observed in minocycline and pregnenolone augmentation trials. The safety of five included anti-inflammatory drugs was also confirmed with the absence of additional side effects.

Psychopathology

The two NSAIDs included in our analysis greatly differed in their effects in PANSS score changes. The efficacy of aspirin was evident with a substantial ES (g = 1.31, 1.69, 1.01, each), but celecoxib augmentation trials did not show significant therapeutic potential. This finding is consistent with a previous review of NSAID augmentation (Nitta et al., 2013). This finding was largely driven by a patient subgroup of high baseline PANSS scores. Both adjuvant estrogen and raloxifene were superior to antipsychotics alone in ameliorating the psychopathology of schizophrenia, which is in line with a recent meta-analysis of sex hormones (Heringa et al., 2015). Nonetheless, the estimated ESs of raloxifene did not remain significant after the trim-and-fill procedure. Pregnenolone, an estrogen precursor, was also significantly beneficial in reducing total PANSS symptom scores, with a moderate ES. A previous meta-analysis including a sample of a 10-week intervention trial with patients having minimal baseline symptoms (Marx et al., 2014) did not yield significant effects for pregnenolone (Heringa et al., 2015). Considering the high baseline PANSS total scores (⩾61) of samples included in our analysis, this discrepancy might have resulted from the selective benefits of pregnenolone dependent on the clinical status of patients.

Omega-3 exerts anti-inflammatory properties by modulating eicosanoid metabolism and cytokine production (Calder, 2015; Mori and Beilin, 2004; Simopoulos, 2002). Nonetheless, the treatment efficacy of omega-3 fatty acids was not significant in the present review. However, the results of meta-regression suggested the therapeutic potential of omega-3 for participants with a short duration of illness. This result is mirrored by the subgroup analysis of a previous review, which revealed that the therapeutic action of omega-3 might be specifically beneficial for patients in the prodromal phase or FEP (Chen et al., 2015). Minocycline effectively reduced negative symptoms, but the high heterogeneity limited the reliability of the results. A recent meta-analysis of minocycline intervention, including trials with an open-label design or BPRS assessment, found beneficial effects of minocycline in reducing both total and negative symptoms (Xiang et al., 2017). Their subgroup analysis suggested several modulating factors to explain the high heterogeneity, e.g. the types of baseline antipsychotics, age and ethnicity of study participants. However, our analysis of potential moderator variables did not account for the between-study heterogeneity. NAC had significant efficacy in treating total and negative symptoms (g = 0.65 and 0.70, each), which confirms the results of the previous meta-analysis (Sommer et al., 2013), with larger patient samples. Notably, the effect of NAC on total symptom improvement was mainly driven by negative symptom improvement. The single study with erythropoietin did not reveal any significant effects in terms of psychopathology.

Identifying moderator variables for the effectiveness of anti-inflammatory agents would be of value for the design of future clinical trials. According to our meta-analysis, adjuvant anti-inflammatory drugs are recommendable for patients with severe baseline total symptoms, as a strong positive relationship was observed between the efficacy of anti-inflammatory agents and the baseline severity of total symptoms in the meta-regression of overall anti-inflammatory augmentation trials. This association was also replicated at a trend level among omega-3 trials (p = 0.10, n = 11) and was further supported by a trend for a lower baseline antipsychotic dose yielding larger ESs for positive symptom improvement in overall trials using anti-inflammatory agents (p = 0.12, n = 17). Previous literature reviews have mainly suggested that patients in an early disease stage tend to benefit more from anti-inflammatory drugs (Mitra et al., 2017; Müller, 2017). Our results have also verified this finding by showing that the efficacy of omega-3 in positive symptoms was negatively impacted by the illness duration of participants. Taken together, these results suggest that acutely ill patients with schizophrenia would benefit more from anti-inflammatory augmentation.

General functioning

A combined ES revealed a significant improvement in GAF scores, which assess psychological, social and occupational functions (g = 0.22). However, no single agent could be recommended in particular for functional improvement, since subgroup analyses of individual agents did not yield significant effects. The lack of effects on functional improvement could have been influenced by uncontrolled variables such as social support and vocational opportunities (Green et al., 2004).

Cognition

An increased level of systematic inflammation in patients with schizophrenia has been linked to changes in cognition and behaviors (Khandaker et al., 2015). Excessive activation of the immune system causes cognitive dysfunction directly through the effects of peripheral cytokines (Harrison et al., 2014; Yirmiya and Goshen, 2011) and activation of microglia (Doorduin et al., 2009; Van Berckel et al., 2008) or indirectly through modulation of neurotransmitters (Meyer and Feldon, 2009) and neurodegenerative processes (Davis et al., 2012). However, whether anti-inflammatory therapies can mitigate cognitive deterioration in schizophrenia has not been examined with meta-analytic approaches due to a lack of data. In our meta-analysis, the cognitive improvement was significant with minocycline and pregnenolone. Visual learning/memory, attention and executive function were the cognitive domains with significant changes in minocycline or pregnenolone augmentation trials. Although these results are promising in that attention, memory and executive function are the most damaged cognitive domains in schizophrenia (Buoli and Altamura, 2015), some caution seems justified as the cognitive scales used in each trial are quite heterogeneous.

Side effects

We did not find any significant changes in ESRS or AIMS scores with celecoxib, omega-3, estrogen, raloxifene, pregnenolone, minocycline and NAC augmentation. Moreover, no significant adverse effects were observed in clinical trials with other agents. Davunetide 5 or 20 mg resulted in a mild degree of self-reported restlessness (Javitt et al., 2012). Chronic administration of NSAIDs could cause gastric side effects (Hirsh et al., 1992), but the safety of aspirin was increased by the concomitant use of omeprazole, a proton inhibitor, in a previous clinical trial (Laan et al., 2010). Erythropoietin has hematopoietic properties, but thromboembolic effects on blood pressure were not reported in the treatment group (Ehrenreich et al., 2007).

The overall meta-analytic results indicated that a wide range of substances with anti-inflammatory actions could help patients with schizophrenia, thereby supporting the involvement of heightened inflammatory responses in the pathology of schizophrenia. Since the use of anti-inflammatory agents for schizophrenia is still in early stages, further investigation is required for the potential efficacy of other promising anti-inflammatory agents. Sodium benzoate, for instance, has been reported to enhance the antipsychotic effect in patients with chronic schizophrenia (Lane et al., 2013; Lin et al., 2018). As suggested in previous reviews, anti-inflammatory agents might inhibit excessive microglial activation and N-methyl-d-aspartate (NMDA) receptor hypofunction (Leza et al., 2015; Müller et al., 2013), which are associated with neurodegeneration in schizophrenia, by down-regulating the immune response. We cannot be sure if the beneficial effects of each anti-inflammatory agent shown in our meta-analysis are evoked by its anti-inflammatory action. Nonetheless, several clinical trials reported that the benefits of NSAID augmentation were limited to a subgroup with abnormal baseline levels of inflammatory markers, such as low serum TNF-α receptor level (Müller et al., 2004) and low TH1/TH2 cytokine ratio (Laan et al., 2010). These results suggest that the anti-inflammatory agents work by normalizing the immune system of patients. Future trials could specifically target patients with elevated inflammatory status for better treatment efficacy. In addition, studies to predict the therapeutic response of anti-inflammatory agents based on inflammatory markers measured at baseline.

Limitations

This study was subject to several limitations. First, other than primary anti-inflammatory agents, we have also examined broad-acting agents with anti-inflammatory properties. The therapeutic efficacy of those drugs should not be entirely ascribed to anti-inflammatory action. However, the ES of PANSS score change was still significant when the analysis was limited to celecoxib and aspirin. Second, subjects in any phase of illness and baseline symptom were included in our analysis. Meta-regression and subgroup analysis suggested that the variability of baseline clinical characteristics increased the overall heterogeneity of our meta-analytic results. Future studies with more specific patient selection are warranted to control for confounding factors. Finally, we have focused on the most frequently used scales (AIMS, ESRS and GAF) for our secondary outcome measures. Including other scales, Glasgow Antipsychotic Side-effect Scale (GASS), Medication Adherence Rating Scale (MARS) or Personal and Social Performance (PSP), Social and Occupational Functioning Assessment Scale (SOFAS) for instance, could permit a more comprehensive understanding of the effects of anti-inflammatory drugs.

Conclusion

Overall, the quantitative evidence proved that anti-inflammatory adjuvant treatments were favorable to antipsychotic monotherapy in terms of psychopathology and real-life functioning. Three cognitive domains, attention, memory and executive function, were enhanced by adjuvant pregnenolone or minocycline. However, whether these agents derived their treatment efficacy from their anti-inflammatory actions remains unclear. Future clinical trials with anti-inflammatory agents could examine the association between changes in inflammatory markers and symptom improvements to clarify this important issue.

Supplemental Material

supplementary_materials – Supplemental material for Adjunctive use of anti-inflammatory drugs for schizophrenia: A meta-analytic investigation of randomized controlled trials

Supplemental material, supplementary_materials for Adjunctive use of anti-inflammatory drugs for schizophrenia: A meta-analytic investigation of randomized controlled trials by Myeongju Cho, Tae Young Lee, Yoo Bin Kwak, Youngwoo Brian Yoon, Minah Kim and Jun Soo Kwon in Australian & New Zealand Journal of Psychiatry

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

This research was supported by the Brain Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT and Future Planning (Grant Nos. 2017M3C7A1029610 and 2016R1E1A1A02921618).

ORCID iD

Youngwoo Brian Yoon

Supplemental material

Supplemental material for this article is available online.

References

Akhondzadeh

Nejatisafa

Amini

, et al. (2003) Adjunctive estrogen treatment in women with chronic schizophrenia: A double-blind, randomized, and placebo-controlled trial. Progress in Neuro-Psychopharmacology & Biological Psychiatry 27: 1007–1012.

Akhondzadeh

Tabatabaee

Amini

, et al. (2007) Celecoxib as adjunctive therapy in schizophrenia: A double-blind, randomized and placebo-controlled trial. Schizophrenia Research 90: 179–185.

Attari

Mojdeh

Soltani

FAST

, et al. (2017) Aspirin inclusion in antipsychotic treatment on severity of symptoms in schizophrenia: A randimized clinical trial. Iranian Journal of Psychiatry and Behavioral Sciences 11: e5848.

Berk

Copolov

Dean

, et al. (2008) N-acetyl cysteine as a glutathione precursor for schizophrenia—A double-blind, randomized, placebo-controlled trial. Biological Psychiatry 64: 361–368.

Bentsen

Osnes

Refsum

, et al. (2013)randomized placebo-controlled trial of an omega-3 fatty acid and vitamins E+C in schizophrenia. Translational psychiatry 3(12):e335.

Boskovic

Vovk

Koprivsek

, et al. (2016) Vitamin E and essential polyunsaturated fatty acids supplementation in schizophrenia patients treated with haloperidol. Nutritional Neuroscience 19: 156–161.

Brown

Derkits

(2009) Prenatal infection and schizophrenia: A review of epidemiologic and translational studies. American Journal of Psychiatry 167: 261–280.

Buoli

Altamura

(2015) May non-antipsychotic drugs improve cognition of schizophrenia patients? Pharmacopsychiatry 48: 41–50.

Calder

(2015) Marine omega-3 fatty acids and inflammatory processes: Effects, mechanisms and clinical relevance. Biochimica et Biophysica Acta (BBA)—Molecular and Cell Biology of Lipids 1851: 469–484.

10.

Chaudhry

Hallak

Husain

, et al. (2012) Minocycline benefits negative symptoms in early schizophrenia: A randomised double-blind placebo-controlled clinical trial in patients on standard treatment. Journal of Psychopharmacology 26: 1185–1193.

11.

Chen

Chibnall

Nasrallah

(2015) A meta-analysis of placebo-controlled trials of omega-3 fatty acid augmentation in schizophrenia: Possible stage-specific effects. Annals of Clinical Psychiatry 27: 289–296.

12.

Davis

Muniz Terrera

Keage

, et al. (2012) Delirium is a strong risk factor for dementia in the oldest-old: A population-based cohort study. Brain 135: 2809–2816.

13.

Dixon

Lehman

Levine

(1995) Conventional antipsychotic medications for schizophrenia. Schizophrenia Bulletin 21: 567.

14.

Doorduin

de Vries

Willemsen

, et al. (2009) Neuroinflammation in schizophrenia-related psychosis: A PET study. Journal of Nuclear Medicine 50: 1801–1807.

15.

Ehrenreich

Hinze-Selch

Stawicki

, et al. (2007) Improvement of cognitive functions in chronic schizophrenic patients by recombinant human erythropoietin. Molecular Psychiatry 12: 206–220.

16.

Emsley

Myburgh

Oosthuizen

, et al. (2002) Randomized, placebo-controlled study of ethyl-eicosapentaenoic acid as supplemental treatment in schizophrenia. American Journal of Psychiatry 159: 1596–1598.

17.

Emsley

Niehaus

Koen

, et al. (2006) The effects of eicosapentaenoic acid in tardive dyskinesia: A randomized, placebo-controlled trial. Schizophrenia Research 84: 112–120.

18.

Erhart

Marder

Carpenter

(2006) Treatment of schizophrenia negative symptoms: Future prospects. Schizophrenia Bulletin 32: 234–237.

19.

Farokhnia

Azarkolah

Adinehfar

, et al. (2013) N-acetylcysteine as an adjunct to risperidone for treatment of negative symptoms in patients with chronic schizophrenia: A randomized, double-blind, placebo-controlled study. Clinical Neuropharmacology 36: 185–192.

20.

Fenton

Dickerson

Boronow

, et al. (2001) A placebo-controlled trial of omega-3 fatty acid (ethyl eicosapentaenoic acid) supplementation for residual symptoms and cognitive impairment in schizophrenia. American Journal of Psychiatry 158: 2071–2074.

21.

Ghafari

Fararouie

Shirazi

, et al. (2013) Combination of estrogen and antipsychotics in the treatment of women with chronic schizophrenia: A double-blind, randomized, placebo-controlled clinical trial. Clinical Schizophrenia & Related Psychoses 6: 172–176.

22.

Ghanizadeh

Dehbozorgi

OmraniSigaroodi

, et al. (2014) Minocycline as add-on treatment decreases the negative symptoms of schizophrenia; a randomized placebo-controlled clinical trial. Recent Patents on Inflammation & Allergy Drug Discovery 8: 211–215.

23.

Gharabawi

Bossie

Lasser

, et al. (2005) Abnormal Involuntary Movement Scale (AIMS) and Extrapyramidal Symptom Rating Scale (ESRS): Cross-scale comparison in assessing tardive dyskinesia. Schizophrenia Research 77: 119–128.

24.

Green

Kern

Heaton

(2004) Longitudinal studies of cognition and functional outcome in schizophrenia: Implications for MATRICS. Schizophrenia Research 72: 41–51.

25.

Hanson

Gottesman

(2005) Theories of schizophrenia: A genetic-inflammatory-vascular synthesis. BMC Medical Genetics 6: 7.

26.

Harrison

Doeller

Voon

, et al. (2014) Peripheral inflammation acutely impairs human spatial memory via actions on medial temporal lobe glucose metabolism. Biological Psychiatry 76: 585–593.

27.

Heringa

Begemann

Goverde

, et al. (2015) Sex hormones and oxytocin augmentation strategies in schizophrenia: A quantitative review. Schizophrenia Research 168: 603–613.

28.

Hirsh

Dalen

Fuster

, et al. (1992) Aspirin and other platelet-active drugs: The relationship between dose, effectiveness, and side effects. Chest 102: 327S–327S.

29.

Jamilian

Solhi

Jamilian

(2014) Randomized, placebo-controlled clinical trial of omega-3 as supplemental treatment in schizophrenia. Global Journal of Health Science 6: 103–108.

30.

Janicak

Marder

Tandon

, et al. (2014) Schizophrenia: Recent Advances in Diagnosis and Treatment. New York: Springer Science+Business Media.

31.

Javitt

Buchanan

Keefe

, et al. (2012) Effect of the neuroprotective peptide davunetide (AL-108) on cognition and functional capacity in schizophrenia. Schizophrenia Research 136: 25–31.

32.

Weickert

Lenroot

, et al. (2016) Adjunctive selective estrogen receptor modulator increases neural activity in the hippocampus and inferior frontal gyrus during emotional face recognition in schizophrenia. Translational Psychiatry 6: e795.

33.

Keefe

Harvey

Goldberg

, et al. (2008) Norms and standardization of the Brief Assessment of Cognition in Schizophrenia (BACS). Schizophrenia Research 102: 108–115.

34.

Keller

Kum

Wehring

, et al. (2013) A review of anti-inflammatory agents for symptoms of schizophrenia. Journal of Psychopharmacology 27: 337–342.

35.

Khandaker

Zimbron

Lewis

, et al. (2013) Prenatal maternal infection, neurodevelopment and adult schizophrenia: A systematic review of population-based studies. Psychological Medicine 43: 239–257.

36.

Khandaker

Cousins

Deakin

, et al. (2015) Inflammation and immunity in schizophrenia: Implications for pathophysiology and treatment. The Lancet Psychiatry 2: 258–270.

37.

Khodaie-Ardakani

Khosravi

Zarinfard

, et al. (2015) A placebo-controlled study of raloxifene added to risperidone in men with chronic schizophrenia. Acta Medica Iranica 53: 337–345.

38.

Khodaie-Ardakani

Mirshafiee

Farokhnia

, et al. (2014) Minocycline add-on to risperidone for treatment of negative symptoms in patients with stable schizophrenia: Randomized double-blind placebo-controlled study. Psychiatry Research 215: 540–546.

39.

Kianimehr

Fatehi

Hashempoor

, et al. (2014) Raloxifene adjunctive therapy for postmenopausal women suffering from chronic schizophrenia: A randomized double-blind and placebo controlled trial. Daru: Journal of Faculty of Pharmacy 22: 55.

40.

Joe

Cho

, et al. (2006) Effect of hormone replacement therapy on cognitive function in women with chronic schizophrenia. International journal of psychiatry in clinical practice 1;10(2):97–104.

41.

Kreinin

Bawakny

Ritsner

(2014) Adjunctive pregnenolone ameliorates the cognitive deficits in recent-onset schizophrenia: An 8-week, randomized, double-blind, placebo-controlled trial. Clinical Schizophrenia & Related Psychoses 10: 201–210.

42.

Kucharska-Pietura

Mortimer

(2013) Can antipsychotics improve social cognition in patients with schizophrenia? CNS Drugs 27: 335–343.

43.

Kulkarni

de Castella

Fitzgerald

, et al. (2008) Estrogen in severe mental illness: A potential new treatment approach. Archives of General Psychiatry 65: 955–960.

44.

Kulkarni

de Castella

Headey

, et al. (2011) Estrogens and men with schizophrenia: Is there a case for adjunctive therapy? Schizophrenia Research 125: 278–283.

45.

Kulkarni

Gavrilidis

Gwini

, et al. (2016) Effect of adjunctive raloxifene therapy on severity of refractory schizophrenia in women: A randomized clinical trial. JAMA Psychiatry 73: 947–954.

46.

Kulkarni

Gavrilidis

Wang

, et al. (2015) Estradiol for treatment-resistant schizophrenia: A large-scale randomized-controlled trial in women of child-bearing age. Molecular Psychiatry 20: 695–702.

47.

Kulkarni

Gurvich

Lee

, et al. (2010) Piloting the effective therapeutic dose of adjunctive selective estrogen receptor modulator treatment in postmenopausal women with schizophrenia. Psychoneuroendocrinology 35: 1142–1147.

48.

Kulkarni

Riedel

de Castella

, et al. (2001) Estrogen—A potential treatment for schizophrenia. Schizophrenia Research 48: 137–144.

49.

Laan

Grobbee

Selten

, et al. (2010) Adjuvant aspirin therapy reduces symptoms of schizophrenia spectrum disorders: Results from a randomized, double-blind, placebo-controlled trial. Journal of Clinical Psychiatry 71: 520–527.

50.

Lane

H-Y

Lin

C-H

Green

, et al. (2013) Add-on treatment of benzoate for schizophrenia: A randomized, double-blind, placebo-controlled trial of D-amino acid oxidase inhibitor. JAMA Psychiatry 70: 1267–1275.

51.

Levkovitz

Mendlovich

Riwkes

, et al. (2010) A double-blind, randomized study of minocycline for the treatment of negative and cognitive symptoms in early-phase schizophrenia. Journal of Clinical Psychiatry 71: 138–149.

52.

Leza

García-Bueno

Bioque

, et al. (2015) Inflammation in schizophrenia: A question of balance. Neuroscience & Biobehavioral Reviews 55: 612–626.

53.

Lin

C-H

Lin

C-H

Chang

Y-C

, et al. (2018) Sodium benzoate, a D-amino acid oxidase inhibitor, added to clozapine for the treatment of schizophrenia: A randomized, double-blind, placebo-controlled trial. Biological Psychiatry 84: 422–432.

54.

Lipsey

Wilson

(2001) Practical Meta-Analysis (Applied Social Research Methods Series), Vol. 49. Thousand Oaks, CA: SAGE.

55.

Liu

Guo

, et al. (2014) Minocycline supplementation for treatment of negative symptoms in early-phase schizophrenia: A double blind, randomized, controlled trial. Schizophrenia Research 153: 169–176.

56.

Manteghiy

Shakeri

Koohestani

Salari

(2008) Beneficial antipsychotic effects of omega-3 fatty acids add-on therapy for the pharmacological management of patients with schizophrenia. Iran J Psychiatry Behav Sci. 2:35–40.

57.

Marx

Keefe

Buchanan

, et al. (2009) Proof-of-concept trial with the neurosteroid pregnenolone targeting cognitive and negative symptoms in schizophrenia. Neuropsychopharmacology 34: 1885–1903.

58.

Marx

Lee

Subramaniam

, et al. (2014) Proof-of-concept randomized controlled trial of pregnenolone in schizophrenia. Psychopharmacology 231: 3647–3662.

59.

Meyer

Feldon

(2009) Neural basis of psychosis-related behaviour in the infection model of schizophrenia. Behavioural Brain Research 204: 322–334.

60.

Meyer

Schwarz

Müller

(2011) Inflammatory processes in schizophrenia: A promising neuroimmunological target for the treatment of negative/cognitive symptoms and beyond. Pharmacology & Therapeutics 132: 96–110.

61.

Miller

Buckley

Seabolt

, et al. (2011) Meta-analysis of cytokine alterations in schizophrenia: Clinical status and antipsychotic effects. Biological Psychiatry 70: 663–671.

62.

Miller

Culpepper

Rapaport

, et al. (2013) Prenatal inflammation and neurodevelopment in schizophrenia: A review of human studies. Progress in Neuro-Psychopharmacology and Biological Psychiatry 42: 92–100.

63.

Mitra

Natarajan

Ziedonis

, et al. (2017) Antioxidant and anti-inflammatory nutrient status, supplementation, and mechanisms in patients with schizophrenia. Progress in Neuro-Psychopharmacology and Biological Psychiatry 78: 1–11.

64.

Miyamoto

Miyake

Jarskog

, et al. (2012) Pharmacological treatment of schizophrenia: A critical review of the pharmacology and clinical effects of current and future therapeutic agents. Molecular Psychiatry 17: 1206–1227.

65.

Monji

Kato

Kanba

(2009) Cytokines and schizophrenia: Microglia hypothesis of schizophrenia. Psychiatry and Clinical Neurosciences 63: 257–265.

66.

Mori

Beilin

(2004) Omega-3 fatty acids and inflammation. Current Atherosclerosis Reports 6: 461–467.

67.

Müller

(2013) The role of anti-inflammatory treatment in psychiatric disorders. Psychiatria Danubina 25: 292–298.

68.

Müller

(2017) Immunological aspects of the treatment of depression and schizophrenia. Dialogues in Clinical Neuroscience 19: 55–63.

69.

Müller

Krause

Dehning

, et al. (2010) Celecoxib treatment in an early stage of schizophrenia: Results of a randomized, double-blind, placebo-controlled trial of celecoxib augmentation of amisulpride treatment. Schizophrenia Research 121: 118–124.

70.

Müller

Myint

A-M

Krause

, et al. (2013) Anti-inflammatory treatment in schizophrenia. Progress in Neuro-Psychopharmacology and Biological Psychiatry 42: 146–153.

71.

Müller

Riedel

Scheppach

, et al. (2002) Beneficial antipsychotic effects of celecoxib add-on therapy compared to risperidone alone in schizophrenia. American Journal of Psychiatry 159: 1029–1034.

72.

Müller

Riedel

Schwarz

, et al. (2005) Clinical effects of COX-2 inhibitors on cognition in schizophrenia. European Archives of Psychiatry and Clinical Neuroscience 255: 149–151.

73.

Müller

Ulmschneider

Scheppach

, et al. (2004) COX-2 inhibition as a treatment approach in schizophrenia: Immunological considerations and clinical effects of celecoxib add-on therapy. European Archives of Psychiatry and Clinical Neuroscience 254: 14–22.

74.

Munetz

Benjamin

(1988) How to examine patients using the Abnormal Involuntary Movement Scale. Hospital & Community Psychiatry 39: 1172–1177.

75.

Najjar

Pearlman

(2015) Neuroinflammation and white matter pathology in schizophrenia: Systematic review. Schizophrenia Research 161: 102–112.

76.

Nitta

Kishimoto

Müller

, et al. (2013) Adjunctive use of nonsteroidal anti-inflammatory drugs for schizophrenia: A meta-analytic investigation of randomized controlled trials. Schizophrenia Bulletin 39: 1230–1241.

77.

Nuechterlein

Green

Kern

, et al. (2008) The MATRICS Consensus Cognitive Battery, part 1: Test selection, reliability, and validity. American Journal of Psychiatry 165: 203–213.

78.

Pawelczyk

Grancow-Grabka

Kotlicka-Antczak

, et al. (2016) A randomized controlled study of the efficacy of six-month supplementation with concentrated fish oil rich in omega-3 polyunsaturated fatty acids in first episode schizophrenia. Journal of Psychiatric Research 73: 34–44.

79.

Peet

Horrobin

(2002) A dose-ranging exploratory study of the effects of ethyl-eicosapentaenoate in patients with persistent schizophrenic symptoms. Journal of Psychiatric Research 36: 7–18.

80.

Peet

Brind

Ramchand

, et al. (2001) Two double-blind placebo-controlled pilot studies of eicosapentaenoic acid in the treatment of schizophrenia. Schizophrenia Research 49: 243–251.

81.

Qiao

Mei

Han

, et al. (2017) Effects of Omega-3 in the treatment of violent schizophrenia patients. Schizophrenia Research 195: 283–285.

82.

Ransohoff

Schafer

Vincent

, et al. (2015) Neuroinflammation: Ways in which the immune system affects the brain. Neurotherapeutics 12: 896–909.

83.

Rapaport

Delrahim

Bresee

, et al. (2005) Celecoxib augmentation of continuously ill patients with schizophrenia. Biological Psychiatry 57: 1594–1596.

84.

Ripke

Neale

Corvin

, et al. (2014) Biological insights from 108 schizophrenia-associated genetic loci. Nature 511: 421–427.

85.

Ritsner

Bawakny

Kreinin

(2014) Pregnenolone treatment reduces severity of negative symptoms in recent-onset schizophrenia: An 8-week, double-blind, randomized add-on two-center trial. Psychiatry and Clinical Neurosciences 68: 432–440.

86.

Ritsner

Gibel

Shleifer

, et al. (2010) Pregnenolone and dehydroepiandrosterone as an adjunctive treatment in schizophrenia and schizoaffective disorder: An 8-week, double-blind, randomized, controlled, 2-center, parallel-group trial. Journal of Clinical Psychiatry 71: 1351–1362.

87.

Rössler

Salize

van Os

, et al. (2005) Size of burden of schizophrenia and psychotic disorders. European Neuropsychopharmacology 15: 399–409.

88.

Schulz

Altman

Moher

(2010) CONSORT 2010 statement: Updated guidelines for reporting parallel group randomised trials. BMJ 340: c332.

89.

Seibert

Zhang

Leahy

, et al. (1994) Pharmacological and biochemical demonstration of the role of cyclooxygenase 2 in inflammation and pain. Proceedings of the National Academy of Sciences of the United States of America 91: 12013–12017.

90.

Simopoulos

(2002) Omega-3 fatty acids in inflammation and autoimmune diseases. Journal of the American College of Nutrition 21: 495–505.

91.

Sommer

de Witte

Begemann

, et al. (2012) Nonsteroidal anti-inflammatory drugs in schizophrenia: Ready for practice or a good start? A meta-analysis. Journal of Clinical Psychiatry 73: 414–419.

92.

Sommer

van Westrhenen

Begemann

, et al. (2013) Efficacy of anti-inflammatory agents to improve symptoms in patients with schizophrenia: An update. Schizophrenia Bulletin 40: 181–191.

93.

Torrey

Peterson

(1973) Slow and latent viruses in schizophrenia. The Lancet 302: 22–24.

94.

Upthegrove

Manzanares-Teson

Barnes

(2014) Cytokine function in medication-naive first episode psychosis: A systematic review and meta-analysis. Schizophrenia Research 155: 101–108.

95.

Usall

Huerta-Ramos

Iniesta

, et al. (2011) Raloxifene as an adjunctive treatment for postmenopausal women with schizophrenia: A double-blind, randomized, placebo-controlled trial. Journal of Clinical Psychiatry 72: 1552–1557.

96.

Usall

Huerta-Ramos

Labad

, et al. (2016) Raloxifene as an adjunctive treatment for postmenopausal women with schizophrenia: A 24-week double-blind, randomized, parallel, placebo-controlled trial. Schizophrenia Bulletin 42: 309–317.

97.

Van Berckel

Bossong

Boellaard

, et al. (2008) Microglia activation in recent-onset schizophrenia: A quantitative (R)-[11C]PK11195 positron emission tomography study. Biological Psychiatry 64: 820–822.

98.

Weickert

Weinberg

Lenroot

, et al. (2015) Adjunctive raloxifene treatment improves attention and memory in men and women with schizophrenia. Molecular Psychiatry 20: 685–694.

99.

Weiser

Levi

Burshtein

, et al. (2017) Raloxifene plus antipsychotics versus placebo plus antipsychotics in severely ill decompensated postmenopausal women with schizophrenia or schizoaffective disorder: A randomized controlled trial. Journal of Clinical Psychiatry 78: e758–e765.

100.

Wiersma

Nienhuis

Slooff

, et al. (1998) Natural course of schizophrenic disorders: A 15-year followup of a Dutch incidence cohort. Schizophrenia Bulletin 24: 75–85.

101.

Wood

Cocchi

Proffitt

, et al. (2010) Neuroprotective effects of ethyl-eicosapentaenoic acid in first episode psychosis: A longitudinal T2 relaxometry pilot study. Psychiatry Research—Neuroimaging 182: 180–182.

102.

Xiang

Y-Q

Zheng

Wang

S-B

, et al. (2017) Adjunctive minocycline for schizophrenia: A meta-analysis of randomized controlled trials. European Neuropsychopharmacology 27: 8–18.

103.

Yirmiya

Goshen

(2011) Immune modulation of learning, memory, neural plasticity and neurogenesis. Brain, Behavior, and Immunity 25: 181–213.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.30 MB