Meta‐Analysis Report: Relationship Between Glutathione Levels and Symptom Severity in Schizophrenia Spectrum Disorders
Danilo Assis Pereira
2025-02-05
Abstract
Glutathione (GSH) is the principal antioxidant safeguarding neural tissue from oxidative damage. Schizophrenia spectrum disorders (SSDs) are characterized not only by positive symptoms but also by persistent negative symptoms and cognitive deficits that are closely tied to functional outcomes. Although both brain imaging studies (using magnetic resonance spectroscopy [MRS]) and blood-based biochemical assays have investigated GSH levels—including total (GSHt), reduced (GSHr), and oxidized (GSSG) forms—the findings have been inconsistent. This systematic review and meta‐analysis synthesized data from studies reporting correlations between GSH levels and clinical parameters (symptom severity, cognitive performance, and functional capacity) in SSD populations. Random‐effects models with Fisher’s Z transformation were used to pool effect sizes. Overall, while MRS‐derived brain GSH measures did not consistently correlate with clinical symptomatology, blood GSHt levels were significantly negatively associated with total and negative symptoms and positively related to cognitive and executive functions. These results suggest that peripheral GSHt may serve as a useful biomarker for SSD severity.
Key words: Glutathione, Schizophrenia Spectrum Disorders, Meta‐analysis, Magnetic Resonance Spectroscopy, Biomarkers, Symptom Severity, Cognitive Function.
Introduction
Schizophrenia spectrum disorders (SSDs) are complex, chronic conditions affecting approximately 1% of the global population. While positive symptoms (e.g., hallucinations and delusions) are often ameliorated with current treatments, negative symptoms and cognitive impairments remain largely refractory and are strongly linked to poor functional outcomes. Among the biological mechanisms implicated in SSD, oxidative stress has gained attention because of its role in neuronal damage. Glutathione (GSH), a tripeptide composed of glutamate, cysteine, and glycine, is the major endogenous antioxidant in the brain and peripheral tissues. It exists in different forms—total (GSHt), reduced (GSHr), and oxidized (GSSG)—and is thought to be involved in modulating neurotransmission (including glutamatergic signaling) and cellular redox balance.
Despite a substantial number of studies using MRS to quantify brain GSH and biochemical assays to measure blood GSH levels, the literature remains inconclusive regarding their relationship with symptom severity, cognitive performance, and overall functioning in SSDs. The objective of this meta‐analysis is to integrate findings from both imaging and peripheral studies in order to clarify whether GSH levels are associated with clinical outcomes and could potentially serve as biomarkers of symptom burden in SSD populations.
Methods
A systematic review and meta‐analysis were conducted in line with PRISMA guidelines. Databases including PubMed, Scopus, Web of Science, Medline, and PsycINFO were searched for studies that: - Enrolled SSD patients (and healthy controls, when applicable); - Reported measurements of brain GSH (via MRS) and/or blood GSH (GSHt, GSHr, GSSG); - Assessed clinical outcomes such as total, positive, or negative symptom severity, cognitive performance, and functional capacity; - Provided sufficient data to compute standardized effect sizes.
Studies with overlapping samples, subjects outside the age range 16–65 years, or those evaluating non‐relevant oxidative stress markers were excluded. Data extraction included sample size, correlation coefficients between GSH levels and clinical measures, measurement modality, and study identifiers.
Effect sizes were computed after transforming correlation
coefficients into Fisher’s Z values. Variances were derived from sample
sizes (using 1/[n – 3]). Random‐effects models (using REML estimation)
were fitted using the metafor and netmeta
packages in R. Subgroup analyses (by measurement type) and
meta‐regressions (testing the moderating effects of “Measurement” and
“Outcome”) were also performed.
Data Preparation and Analysis Code
Below, the R code loads the required libraries, prepares the data (including separate data frames for imaging and blood studies), computes Fisher’s Z values, and creates a study identifier.
# ============================================================
# 1. LOAD LIBRARIES
# ============================================================
# Uncomment and run the following line if installation is needed:
# install.packages(c("metafor", "netmeta", "UpSetR", "corrplot", "circlize", "ComplexHeatmap", "tidyr", "dplyr", "grid"))
library(metafor)
library(netmeta)
library(UpSetR)
library(corrplot)
library(circlize)
library(ComplexHeatmap)
library(tidyr)
library(dplyr)
library(grid) # for grid.text, grid.rect, etc.
# ============================================================
# 2. PREPARE DATA: CREATE DATA FRAMES FOR EACH SUBGROUP
# ============================================================
# --- Imaging Data (MRS studies) ---
imaging_positive <- data.frame(
Authors = c("Matsuzawa et al. 2008", "Reyes-Madrigal et al. 2019", "Iwata et al. 2021",
"Coughlin et al. 2021", "Lesh et al. 2021"),
Correlation = c(-0.43, 0.96, -0.08, 0.14, -0.266),
SampleSize = c(20, 10, 67, 16, 33),
Measurement = "Imaging",
Outcome = "Positive",
stringsAsFactors = FALSE
)
imaging_negative <- data.frame(
Authors = c("Matsuzawa et al. 2008", "Reyes-Madrigal et al. 2019", "Iwata et al. 2021",
"Coughlin et al. 2021", "Lesh et al. 2021", "Ravanfar et al. 2022"),
Correlation = c(-0.6, 0.36, 0.15, 0.21, -0.01, -0.348),
SampleSize = c(20, 10, 67, 16, 33, 12),
Measurement = "Imaging",
Outcome = "Negative",
stringsAsFactors = FALSE
)
imaging_general <- data.frame(
Authors = c("Reyes-Madrigal et al. 2019", "Iwata et al. 2021"),
Correlation = c(0.14, -0.15),
SampleSize = c(10, 67),
Measurement = "Imaging",
Outcome = "General",
stringsAsFactors = FALSE
)
imaging_total <- data.frame(
Authors = c("Matsuzawa et al. 2008", "Reyes-Madrigal et al. 2019", "Iwata et al. 2021",
"Lesh et al. 2021", "Ravanfar et al. 2022"),
Correlation = c(-0.41, 0.42, -0.08, -0.293, -0.286),
SampleSize = c(20, 10, 67, 33, 12),
Measurement = "Imaging",
Outcome = "Total",
stringsAsFactors = FALSE
)
imaging_ideational_fluency <- data.frame(
Authors = c("Matsuzawa et al. 2008", "Coughlin et al. 2021"),
Correlation = c(0.21, 0.61),
SampleSize = c(36, 16),
Measurement = "Imaging",
Outcome = "Ideational_Fluency",
stringsAsFactors = FALSE
)
imaging_processing_speed <- data.frame(
Authors = c("Matsuzawa et al. 2008", "Coughlin et al. 2021"),
Correlation = c(-0.14, 0.26),
SampleSize = c(36, 16),
Measurement = "Imaging",
Outcome = "Processing_Speed",
stringsAsFactors = FALSE
)
imaging_verbal_memory <- data.frame(
Authors = c("Matsuzawa et al. 2008", "Coughlin et al. 2021"),
Correlation = c(0.18, 0.12),
SampleSize = c(36, 16),
Measurement = "Imaging",
Outcome = "Verbal_Memory",
stringsAsFactors = FALSE
)
imaging_functioning <- data.frame(
Authors = c("Lesh et al. 2021", "Mackinley et al. 2022", "Ravanfar et al. 2022"),
Correlation = c(0.185, 0.04, 0.452),
SampleSize = c(33, 53, 12),
Measurement = "Imaging",
Outcome = "Functioning",
stringsAsFactors = FALSE
)
# --- GSHt (Total Glutathione in blood) Data ---
gsht_positive <- data.frame(
Authors = c("Raffa et al. 2011", "Tsai et al. 2013", "Nucifora et al. 2017", "Hendouei et al. 2018",
"Hendouei et al. 2018*", "Hendouei et al. 2018**", "Chien et al. 2020", "Chien et al. 2020*",
"Coughlin et al. 2021", "Gares-Caballer et al. 2022", "Lin et al. 2023 (N=92)", "Lin et al. 2023 (N=219)",
"Fathy et al. 2015", "Kizilpinar et al. 2023"),
Correlation = c(0.5, -0.304, -0.359, -0.1, -0.2, 0.2, 0.03, 0.22, -0.21, -0.06, 0.078, 0.071, 0.316, -0.139),
SampleSize = c(23, 41, 51, 34, 34, 32, 43, 19, 24, 30, 92, 219, 30, 26),
Measurement = "GSHt",
Outcome = "Positive",
stringsAsFactors = FALSE
)
gsht_negative <- data.frame(
Authors = c("Raffa et al. 2011", "Tsai et al. 2013", "Nucifora et al. 2017", "Hendouei et al. 2018",
"Hendouei et al. 2018*", "Hendouei et al. 2018**", "Chien et al. 2020", "Chien et al. 2020*",
"Coughlin et al. 2021", "Gares-Caballer et al. 2022", "Lin et al. 2023 (N=92)", "Lin et al. 2023 (N=219)",
"Fathy et al. 2015", "Kizilpinar et al. 2023"),
Correlation = c(-0.02, -0.349, -0.203, 0.07, -0.1, -0.1, 0, -0.17, -0.06, 0.01, -0.027, -0.055, -0.805, -0.038),
SampleSize = c(23, 41, 51, 34, 34, 32, 43, 19, 24, 30, 92, 219, 30, 26),
Measurement = "GSHt",
Outcome = "Negative",
stringsAsFactors = FALSE
)
gsht_general <- data.frame(
Authors = c("Nucifora et al. 2017", "Hendouei et al. 2018", "Hendouei et al. 2018*", "Hendouei et al. 2018**",
"Gares-Caballer et al. 2022", "Lin et al. 2023 (N=92)", "Lin et al. 2023 (N=219)", "Kizilpinar et al. 2023"),
Correlation = c(-0.262, -0.1, -0.1, 0.2, 0.001, 0.145, -0.099, 0.037),
SampleSize = c(51, 34, 34, 32, 30, 92, 219, 26),
Measurement = "GSHt",
Outcome = "General",
stringsAsFactors = FALSE
)
gsht_total <- data.frame(
Authors = c("Tuncel et al. 2015", "Tsai et al. 2013", "Nucifora et al. 2017", "Hendouei et al. 2018",
"Hendouei et al. 2018*", "Hendouei et al. 2018**", "Lin et al. 2023 (N=92)", "Lin et al. 2023 (N=219)",
"Kizilpinar et al. 2023"),
Correlation = c(-0.106, -0.413, -0.311, -0.1, -0.1, 0.1, 0.068, -0.047, 0.016),
SampleSize = c(18, 41, 51, 34, 34, 32, 92, 219, 26),
Measurement = "GSHt",
Outcome = "Total",
stringsAsFactors = FALSE
)
gsht_executive <- data.frame(
Authors = c("Martinez-Cengotitobengoa et al. 2012", "Gonzalez-Liencres et al. 2014", "Coughlin et al. 2021", "Gares-Caballer et al. 2022"),
Correlation = c(0.072, 0.171, 0.45, 0.4),
SampleSize = c(28, 41, 24, 30),
Measurement = "GSHt",
Outcome = "Executive",
stringsAsFactors = FALSE
)
gsht_global_cog <- data.frame(
Authors = c("Nucifora et al. 2017", "Coughlin et al. 2021", "Gares-Caballer et al. 2022"),
Correlation = c(0.245, 0.57, 0.34),
SampleSize = c(51, 24, 30),
Measurement = "GSHt",
Outcome = "Global_Cognitive_Score",
stringsAsFactors = FALSE
)
gsht_processing_speed <- data.frame(
Authors = c("Gonzalez-Liencres et al. 2014", "Coughlin et al. 2021", "Gares-Caballer et al. 2022",
"Lin et al. 2023 (N=92)", "Lin et al. 2023 (N=219)"),
Correlation = c(-0.119, 0.41, 0.21, -0.172, 0.045),
SampleSize = c(41, 24, 30, 92, 219),
Measurement = "GSHt",
Outcome = "Processing_Speed",
stringsAsFactors = FALSE
)
gsht_cognitive_flex <- data.frame(
Authors = c("Gonzalez-Liencres et al. 2014", "Coughlin et al. 2021", "Lin et al. 2023 (N=92)", "Lin et al. 2023 (N=219)"),
Correlation = c(0.2, -0.06, -0.059, 0.054),
SampleSize = c(41, 24, 92, 219),
Measurement = "GSHt",
Outcome = "Cognitive_Flexibility",
stringsAsFactors = FALSE
)
gsht_working_memory <- data.frame(
Authors = c("Lin et al. 2023 (N=92)", "Lin et al. 2023 (N=219)"),
Correlation = c(-0.094, -0.027),
SampleSize = c(92, 219),
Measurement = "GSHt",
Outcome = "Working_Memory",
stringsAsFactors = FALSE
)
gsht_cgi <- data.frame(
Authors = c("Raffa et al. 2009", "Gares-Caballer et al. 2022"),
Correlation = c(-0.28, -0.08),
SampleSize = c(88, 30),
Measurement = "GSHt",
Outcome = "CGI",
stringsAsFactors = FALSE
)
gsht_functioning <- data.frame(
Authors = c("Lin et al. 2023 (N=92)", "Lin et al. 2023 (N=219)"),
Correlation = c(-0.107, 0.076),
SampleSize = c(92, 219),
Measurement = "GSHt",
Outcome = "Functioning",
stringsAsFactors = FALSE
)
# --- GSHr (Reduced Glutathione in blood) Data ---
gshr_positive <- data.frame(
Authors = c("Raffa et al. 2011", "Guidara et al. 2020", "Cruz et al. 2021", "Piatoikina et al. 2023", "Altuntas et al. 2000"),
Correlation = c(0.51, -0.147, 0.082, 0.11, -0.18),
SampleSize = c(23, 66, 85, 125, 48),
Measurement = "GSHr",
Outcome = "Positive",
stringsAsFactors = FALSE
)
gshr_negative <- data.frame(
Authors = c("Raffa et al. 2011", "Guidara et al. 2020", "Cruz et al. 2021", "Wiedlocha et al. 2023", "Piatoikina et al. 2023"),
Correlation = c(-0.05, -0.011, 0.036, -0.413, -0.02),
SampleSize = c(23, 66, 85, 82, 125),
Measurement = "GSHr",
Outcome = "Negative",
stringsAsFactors = FALSE
)
gshr_general <- data.frame(
Authors = c("Guidara et al. 2020", "Piatoikina et al. 2023"),
Correlation = c(-0.156, 0.01),
SampleSize = c(66, 125),
Measurement = "GSHr",
Outcome = "General",
stringsAsFactors = FALSE
)
gshr_total <- data.frame(
Authors = c("Guidara et al. 2020", "Piatoikina et al. 2023", "Altuntas et al. 2000"),
Correlation = c(-0.155, 0.03, -0.08),
SampleSize = c(66, 125, 48),
Measurement = "GSHr",
Outcome = "Total",
stringsAsFactors = FALSE
)
gshr_working_memory <- data.frame(
Authors = c("Cruz et al. 2021", "Piatoikina et al. 2021"),
Correlation = c(-0.041, -0.003),
SampleSize = c(85, 125),
Measurement = "GSHr",
Outcome = "Working_Memory",
stringsAsFactors = FALSE
)
gshr_global_cog <- data.frame(
Authors = c("Guidara et al. 2020", "Cruz et al. 2021"),
Correlation = c(0.118, -0.092),
SampleSize = c(66, 85),
Measurement = "GSHr",
Outcome = "Global_Cognitive_Score",
stringsAsFactors = FALSE
)
gshr_executive <- data.frame(
Authors = c("Cruz et al. 2021", "Piatoikina et al. 2021"),
Correlation = c(-0.114, 0.043),
SampleSize = c(85, 125),
Measurement = "GSHr",
Outcome = "Executive",
stringsAsFactors = FALSE
)
gshr_verbal_memory <- data.frame(
Authors = c("Cruz et al. 2021", "Piatoikina et al. 2021"),
Correlation = c(0, 0.02),
SampleSize = c(85, 125),
Measurement = "GSHr",
Outcome = "Verbal_Memory",
stringsAsFactors = FALSE
)
gshr_processing_speed <- data.frame(
Authors = c("Cruz et al. 2021", "Piatoikina et al. 2021"),
Correlation = c(0.038, 0.03),
SampleSize = c(85, 125),
Measurement = "GSHr",
Outcome = "Processing_Speed",
stringsAsFactors = FALSE
)
gshr_cgi <- data.frame(
Authors = c("Raffa et al. 2009", "Ballesteros et al. 2013"),
Correlation = c(-0.32, 2.08), # NOTE: 2.08 is not valid and will be filtered out.
SampleSize = c(88, 54),
Measurement = "GSHr",
Outcome = "CGI",
stringsAsFactors = FALSE
)
# --- GSSG (Oxidized Glutathione in blood) Data ---
gssg_positive <- data.frame(
Authors = c("Raffa et al. 2011", "Tao et al. 2020"),
Correlation = c(0.16, 0.119),
SampleSize = c(23, 90),
Measurement = "GSSG",
Outcome = "Positive",
stringsAsFactors = FALSE
)
gssg_negative <- data.frame(
Authors = c("Raffa et al. 2011", "Tao et al. 2020"),
Correlation = c(0.17, -0.082),
SampleSize = c(23, 90),
Measurement = "GSSG",
Outcome = "Negative",
stringsAsFactors = FALSE
)
# ============================================================
# 3. COMBINE ALL DATA & COMPUTE FISHER'S Z AND VARIANCE
# ============================================================
all_data <- rbind(
imaging_positive,
imaging_negative,
imaging_general,
imaging_total,
imaging_ideational_fluency,
imaging_processing_speed,
imaging_verbal_memory,
imaging_functioning,
gsht_positive,
gsht_negative,
gsht_general,
gsht_total,
gsht_executive,
gsht_global_cog,
gsht_processing_speed,
gsht_cognitive_flex,
gsht_working_memory,
gsht_cgi,
gsht_functioning,
gshr_positive,
gshr_negative,
gshr_general,
gshr_total,
gshr_working_memory,
gshr_global_cog,
gshr_executive,
gshr_verbal_memory,
gshr_processing_speed,
gshr_cgi,
gssg_positive,
gssg_negative
)
# Remove rows with invalid correlations (absolute correlation must be < 1)
all_data <- all_data[abs(all_data$Correlation) < 1, ]
# Compute Fisher's Z transformation and corresponding variance
all_data$FisherZ <- atanh(all_data$Correlation)
all_data$VarFisherZ<- 1/(all_data$SampleSize - 3)
# Create a study identifier based on the Authors column
all_data$Study <- all_data$Authors1. Overall Random‐Effects Meta‐Analysis
To estimate an overall effect size (i.e., the association between GSH levels and clinical measures) and quantify heterogeneity, a random‐effects meta‐analysis was conducted using the Fisher’s Z effect sizes and their variances.
ma_overall <- rma(yi = FisherZ,
vi = VarFisherZ,
data = all_data,
method = "REML")
cat("Overall Random-Effects Meta-Analysis:\n")## Overall Random-Effects Meta-Analysis:print(summary(ma_overall))##
## Random-Effects Model (k = 124; tau^2 estimator: REML)
##
## logLik deviance AIC BIC AICc
## -4.1723 8.3445 12.3445 17.9689 12.4445
##
## tau^2 (estimated amount of total heterogeneity): 0.0162 (SE = 0.0046)
## tau (square root of estimated tau^2 value): 0.1274
## I^2 (total heterogeneity / total variability): 47.91%
## H^2 (total variability / sampling variability): 1.92
##
## Test for Heterogeneity:
## Q(df = 123) = 253.0875, p-val < .0001
##
## Model Results:
##
## estimate se zval pval ci.lb ci.ub
## -0.0192 0.0179 -1.0725 0.2835 -0.0543 0.0159
##
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1Explanation:
This model provides a pooled estimate of the association across all
studies along with heterogeneity statistics (e.g., tau², I²).
2. Meta‐Regression with Moderators
A meta‐regression was performed to assess whether the type of measurement (e.g., Imaging vs. blood GSHt, GSHr, GSSG) and the specific clinical outcome (e.g., Total, Positive, Negative, Executive) moderated the observed associations.
# Ensure moderators are factors
all_data$Measurement <- as.factor(all_data$Measurement)
all_data$Outcome <- as.factor(all_data$Outcome)
ma_meta_reg <- rma(yi = FisherZ,
vi = VarFisherZ,
mods = ~ Measurement + Outcome,
data = all_data,
method = "REML")
cat("Meta-Regression with Measurement and Outcome as Moderators:\n")## Meta-Regression with Measurement and Outcome as Moderators:print(summary(ma_meta_reg))##
## Mixed-Effects Model (k = 124; tau^2 estimator: REML)
##
## logLik deviance AIC BIC AICc
## 1.0395 -2.0790 31.9210 77.5173 38.7210
##
## tau^2 (estimated amount of residual heterogeneity): 0.0137 (SE = 0.0047)
## tau (square root of estimated tau^2 value): 0.1170
## I^2 (residual heterogeneity / unaccounted variability): 41.34%
## H^2 (unaccounted variability / sampling variability): 1.70
## R^2 (amount of heterogeneity accounted for): 15.57%
##
## Test for Residual Heterogeneity:
## QE(df = 108) = 204.0741, p-val < .0001
##
## Test of Moderators (coefficients 2:16):
## QM(df = 15) = 33.0103, p-val = 0.0047
##
## Model Results:
##
## estimate se zval pval ci.lb
## intrcpt -0.2802 0.1040 -2.6947 0.0070 -0.4839
## MeasurementGSHt 0.0275 0.0419 0.6571 0.5111 -0.0546
## MeasurementGSSG 0.1346 0.1035 1.2997 0.1937 -0.0683
## MeasurementImaging 0.0360 0.0605 0.5953 0.5516 -0.0826
## OutcomeCognitive_Flexibility 0.2870 0.1335 2.1502 0.0315 0.0254
## OutcomeExecutive 0.3855 0.1277 3.0193 0.0025 0.1352
## OutcomeFunctioning 0.3074 0.1313 2.3412 0.0192 0.0501
## OutcomeGeneral 0.2118 0.1146 1.8473 0.0647 -0.0129
## OutcomeGlobal_Cognitive_Score 0.4507 0.1322 3.4089 0.0007 0.1916
## OutcomeIdeational_Fluency 0.6193 0.2066 2.9976 0.0027 0.2144
## OutcomeNegative 0.1462 0.1088 1.3439 0.1790 -0.0670
## OutcomePositive 0.2662 0.1090 2.4428 0.0146 0.0526
## OutcomeProcessing_Speed 0.2834 0.1178 2.4054 0.0162 0.0525
## OutcomeTotal 0.1558 0.1123 1.3875 0.1653 -0.0643
## OutcomeVerbal_Memory 0.3239 0.1385 2.3389 0.0193 0.0525
## OutcomeWorking_Memory 0.2266 0.1256 1.8044 0.0712 -0.0195
## ci.ub
## intrcpt -0.0764 **
## MeasurementGSHt 0.1096
## MeasurementGSSG 0.3375
## MeasurementImaging 0.1547
## OutcomeCognitive_Flexibility 0.5485 *
## OutcomeExecutive 0.6357 **
## OutcomeFunctioning 0.5648 *
## OutcomeGeneral 0.4365 .
## OutcomeGlobal_Cognitive_Score 0.7099 ***
## OutcomeIdeational_Fluency 1.0242 **
## OutcomeNegative 0.3594
## OutcomePositive 0.4798 *
## OutcomeProcessing_Speed 0.5142 *
## OutcomeTotal 0.3758
## OutcomeVerbal_Memory 0.5953 *
## OutcomeWorking_Memory 0.4728 .
##
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1Explanation:
This analysis evaluates whether differences in measurement modality or
outcome type explain variability in effect sizes across studies.
3. Subgroup Meta‐Analyses by Measurement Type
To further investigate heterogeneity, separate random‐effects models were fitted for each measurement type.
measurement_levels <- levels(all_data$Measurement)
subgroup_results <- list()
for (m in measurement_levels) {
dat_subset <- subset(all_data, Measurement == m)
ma_sub <- rma(yi = FisherZ,
vi = VarFisherZ,
data = dat_subset,
method = "REML")
subgroup_results[[m]] <- ma_sub
cat("\n--- Meta-Analysis for Measurement:", m, "---\n")
print(summary(ma_sub))
}##
## --- Meta-Analysis for Measurement: GSHr ---
##
## Random-Effects Model (k = 26; tau^2 estimator: REML)
##
## logLik deviance AIC BIC AICc
## 11.6621 -23.3242 -19.3242 -16.8865 -18.7788
##
## tau^2 (estimated amount of total heterogeneity): 0.0061 (SE = 0.0051)
## tau (square root of estimated tau^2 value): 0.0784
## I^2 (total heterogeneity / total variability): 33.66%
## H^2 (total variability / sampling variability): 1.51
##
## Test for Heterogeneity:
## Q(df = 25) = 40.7573, p-val = 0.0243
##
## Model Results:
##
## estimate se zval pval ci.lb ci.ub
## -0.0345 0.0269 -1.2829 0.1995 -0.0872 0.0182
##
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
##
## --- Meta-Analysis for Measurement: GSHt ---
##
## Random-Effects Model (k = 67; tau^2 estimator: REML)
##
## logLik deviance AIC BIC AICc
## 0.3657 -0.7314 3.2686 7.6479 3.4591
##
## tau^2 (estimated amount of total heterogeneity): 0.0225 (SE = 0.0077)
## tau (square root of estimated tau^2 value): 0.1499
## I^2 (total heterogeneity / total variability): 56.97%
## H^2 (total variability / sampling variability): 2.32
##
## Test for Heterogeneity:
## Q(df = 66) = 138.6789, p-val < .0001
##
## Model Results:
##
## estimate se zval pval ci.lb ci.ub
## -0.0226 0.0264 -0.8578 0.3910 -0.0744 0.0291
##
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
##
## --- Meta-Analysis for Measurement: GSSG ---
##
## Random-Effects Model (k = 4; tau^2 estimator: REML)
##
## logLik deviance AIC BIC AICc
## 1.4970 -2.9941 1.0059 -0.7968 13.0059
##
## tau^2 (estimated amount of total heterogeneity): 0.0028 (SE = 0.0181)
## tau (square root of estimated tau^2 value): 0.0533
## I^2 (total heterogeneity / total variability): 11.67%
## H^2 (total variability / sampling variability): 1.13
##
## Test for Heterogeneity:
## Q(df = 3) = 2.4825, p-val = 0.4785
##
## Model Results:
##
## estimate se zval pval ci.lb ci.ub
## 0.0502 0.0751 0.6691 0.5034 -0.0969 0.1974
##
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
##
## --- Meta-Analysis for Measurement: Imaging ---
##
## Random-Effects Model (k = 27; tau^2 estimator: REML)
##
## logLik deviance AIC BIC AICc
## -15.5632 31.1263 35.1263 37.6425 35.6480
##
## tau^2 (estimated amount of total heterogeneity): 0.0881 (SE = 0.0384)
## tau (square root of estimated tau^2 value): 0.2969
## I^2 (total heterogeneity / total variability): 68.97%
## H^2 (total variability / sampling variability): 3.22
##
## Test for Heterogeneity:
## Q(df = 26) = 69.7143, p-val < .0001
##
## Model Results:
##
## estimate se zval pval ci.lb ci.ub
## 0.0392 0.0727 0.5398 0.5893 -0.1033 0.1818
##
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1Explanation:
This step provides pooled effect estimates and heterogeneity statistics
separately for Imaging, GSHt, GSHr, and GSSG studies.
4. Network Meta‐Analysis by Measurement Category
Network meta‐analysis permits simultaneous comparisons across multiple outcomes (treated as “treatments”). For each measurement type, pairwise comparisons among outcomes were generated and analyzed using a random‐effects network meta‐analysis model.
library(netmeta)
network_results <- list()
for (m in measurement_levels) {
dat_net <- subset(all_data, Measurement == m)
# Compute standard errors from the variance of Fisher's Z
dat_net$seTE <- sqrt(dat_net$VarFisherZ)
# Create pairwise comparisons using outcome as the treatment variable
pw <- pairwise(treat = Outcome,
TE = FisherZ,
seTE = seTE,
studlab = Study,
data = dat_net,
sm = "SMD")
# Run the network meta-analysis with a random effects model
net_mod <- netmeta(TE = pw$TE,
seTE = pw$seTE,
treat1 = pw$treat1,
treat2 = pw$treat2,
studlab = pw$studlab,
sm = "SMD",
comb.fixed = FALSE,
comb.random = TRUE)
cat("\n=== Network Meta-Analysis for Measurement:", m, "===\n")
print(summary(net_mod))
# Optional: Plot the network graph
netgraph(net_mod,
plastic = TRUE,
col = "darkblue",
thickness = "se.random")
network_results[[m]] <- net_mod
}##
## === Network Meta-Analysis for Measurement: GSHr ===
## Original data (with adjusted standard errors for multi-arm studies):
##
## treat1 treat2 TE
## Raffa et al. 2011 Negative Positive -0.6128
## Guidara et al. 2020 General Positive -0.0092
## Guidara et al. 2020 Global_Cognitive_Score Positive 0.2666
## Guidara et al. 2020 Negative Positive 0.1371
## Guidara et al. 2020 Positive Total 0.0082
## Cruz et al. 2021 Executive Positive -0.1967
## Cruz et al. 2021 Global_Cognitive_Score Positive -0.1744
## Cruz et al. 2021 Negative Positive -0.0462
## Cruz et al. 2021 Positive Processing_Speed 0.0442
## Cruz et al. 2021 Positive Verbal_Memory 0.0822
## Cruz et al. 2021 Positive Working_Memory 0.1232
## Piatoikina et al. 2023 General Positive -0.1004
## Piatoikina et al. 2023 Negative Positive -0.1304
## Piatoikina et al. 2023 Positive Total 0.0804
## Altuntas et al. 2000 Positive Total -0.1018
## Guidara et al. 2020 General Negative -0.1463
## Guidara et al. 2020 Global_Cognitive_Score Negative 0.1296
## Guidara et al. 2020 Negative Total 0.1453
## Cruz et al. 2021 Executive Negative -0.1505
## Cruz et al. 2021 Global_Cognitive_Score Negative -0.1283
## Cruz et al. 2021 Negative Processing_Speed -0.0020
## Cruz et al. 2021 Negative Verbal_Memory 0.0360
## Cruz et al. 2021 Negative Working_Memory 0.0770
## Piatoikina et al. 2023 General Negative 0.0300
## Piatoikina et al. 2023 Negative Total -0.0500
## Guidara et al. 2020 General Global_Cognitive_Score -0.2758
## Guidara et al. 2020 General Total -0.0010
## Piatoikina et al. 2023 General Total -0.0200
## Guidara et al. 2020 Global_Cognitive_Score Total 0.2748
## Cruz et al. 2021 Executive Working_Memory -0.0735
## Cruz et al. 2021 Global_Cognitive_Score Working_Memory -0.0512
## Cruz et al. 2021 Processing_Speed Working_Memory 0.0790
## Cruz et al. 2021 Verbal_Memory Working_Memory 0.0410
## Piatoikina et al. 2021 Executive Working_Memory 0.0460
## Piatoikina et al. 2021 Processing_Speed Working_Memory 0.0330
## Piatoikina et al. 2021 Verbal_Memory Working_Memory 0.0230
## Cruz et al. 2021 Executive Global_Cognitive_Score -0.0222
## Cruz et al. 2021 Global_Cognitive_Score Processing_Speed -0.1303
## Cruz et al. 2021 Global_Cognitive_Score Verbal_Memory -0.0923
## Cruz et al. 2021 Executive Processing_Speed -0.1525
## Cruz et al. 2021 Executive Verbal_Memory -0.1145
## Piatoikina et al. 2021 Executive Processing_Speed 0.0130
## Piatoikina et al. 2021 Executive Verbal_Memory 0.0230
## Cruz et al. 2021 Processing_Speed Verbal_Memory 0.0380
## Piatoikina et al. 2021 Processing_Speed Verbal_Memory 0.0100
## seTE seTE.adj narms multiarm
## Raffa et al. 2011 0.3162 0.3162 2
## Guidara et al. 2020 0.1782 0.2817 5 *
## Guidara et al. 2020 0.1782 0.2817 5 *
## Guidara et al. 2020 0.1782 0.2817 5 *
## Guidara et al. 2020 0.1782 0.2817 5 *
## Cruz et al. 2021 0.1562 0.2922 7 *
## Cruz et al. 2021 0.1562 0.2922 7 *
## Cruz et al. 2021 0.1562 0.2922 7 *
## Cruz et al. 2021 0.1562 0.2922 7 *
## Cruz et al. 2021 0.1562 0.2922 7 *
## Cruz et al. 2021 0.1562 0.2922 7 *
## Piatoikina et al. 2023 0.1280 0.1811 4 *
## Piatoikina et al. 2023 0.1280 0.1811 4 *
## Piatoikina et al. 2023 0.1280 0.1811 4 *
## Altuntas et al. 2000 0.2108 0.2108 2
## Guidara et al. 2020 0.1782 0.2817 5 *
## Guidara et al. 2020 0.1782 0.2817 5 *
## Guidara et al. 2020 0.1782 0.2817 5 *
## Cruz et al. 2021 0.1562 0.2922 7 *
## Cruz et al. 2021 0.1562 0.2922 7 *
## Cruz et al. 2021 0.1562 0.2922 7 *
## Cruz et al. 2021 0.1562 0.2922 7 *
## Cruz et al. 2021 0.1562 0.2922 7 *
## Piatoikina et al. 2023 0.1280 0.1811 4 *
## Piatoikina et al. 2023 0.1280 0.1811 4 *
## Guidara et al. 2020 0.1782 0.2817 5 *
## Guidara et al. 2020 0.1782 0.2817 5 *
## Piatoikina et al. 2023 0.1280 0.1811 4 *
## Guidara et al. 2020 0.1782 0.2817 5 *
## Cruz et al. 2021 0.1562 0.2922 7 *
## Cruz et al. 2021 0.1562 0.2922 7 *
## Cruz et al. 2021 0.1562 0.2922 7 *
## Cruz et al. 2021 0.1562 0.2922 7 *
## Piatoikina et al. 2021 0.1280 0.1811 4 *
## Piatoikina et al. 2021 0.1280 0.1811 4 *
## Piatoikina et al. 2021 0.1280 0.1811 4 *
## Cruz et al. 2021 0.1562 0.2922 7 *
## Cruz et al. 2021 0.1562 0.2922 7 *
## Cruz et al. 2021 0.1562 0.2922 7 *
## Cruz et al. 2021 0.1562 0.2922 7 *
## Cruz et al. 2021 0.1562 0.2922 7 *
## Piatoikina et al. 2021 0.1280 0.1811 4 *
## Piatoikina et al. 2021 0.1280 0.1811 4 *
## Cruz et al. 2021 0.1562 0.2922 7 *
## Piatoikina et al. 2021 0.1280 0.1811 4 *
##
## Number of treatment arms (by study):
## narms
## Raffa et al. 2011 2
## Altuntas et al. 2000 2
## Piatoikina et al. 2023 4
## Guidara et al. 2020 5
## Cruz et al. 2021 7
## Piatoikina et al. 2021 4
##
## Results (random effects model):
##
## treat1 treat2 SMD
## Raffa et al. 2011 Negative Positive -0.0681
## Guidara et al. 2020 General Positive -0.0994
## Guidara et al. 2020 Global_Cognitive_Score Positive -0.0109
## Guidara et al. 2020 Negative Positive -0.0681
## Guidara et al. 2020 Positive Total 0.0655
## Cruz et al. 2021 Executive Positive -0.0863
## Cruz et al. 2021 Global_Cognitive_Score Positive -0.0109
## Cruz et al. 2021 Negative Positive -0.0681
## Cruz et al. 2021 Positive Processing_Speed 0.0328
## Cruz et al. 2021 Positive Verbal_Memory 0.0541
## Cruz et al. 2021 Positive Working_Memory 0.0843
## Piatoikina et al. 2023 General Positive -0.0994
## Piatoikina et al. 2023 Negative Positive -0.0681
## Piatoikina et al. 2023 Positive Total 0.0655
## Altuntas et al. 2000 Positive Total 0.0655
## Guidara et al. 2020 General Negative -0.0312
## Guidara et al. 2020 Global_Cognitive_Score Negative 0.0573
## Guidara et al. 2020 Negative Total -0.0026
## Cruz et al. 2021 Executive Negative -0.0182
## Cruz et al. 2021 Global_Cognitive_Score Negative 0.0573
## Cruz et al. 2021 Negative Processing_Speed -0.0354
## Cruz et al. 2021 Negative Verbal_Memory -0.0141
## Cruz et al. 2021 Negative Working_Memory 0.0162
## Piatoikina et al. 2023 General Negative -0.0312
## Piatoikina et al. 2023 Negative Total -0.0026
## Guidara et al. 2020 General Global_Cognitive_Score -0.0885
## Guidara et al. 2020 General Total -0.0339
## Piatoikina et al. 2023 General Total -0.0339
## Guidara et al. 2020 Global_Cognitive_Score Total 0.0546
## Cruz et al. 2021 Executive Working_Memory -0.0020
## Cruz et al. 2021 Global_Cognitive_Score Working_Memory 0.0734
## Cruz et al. 2021 Processing_Speed Working_Memory 0.0515
## Cruz et al. 2021 Verbal_Memory Working_Memory 0.0302
## Piatoikina et al. 2021 Executive Working_Memory -0.0020
## Piatoikina et al. 2021 Processing_Speed Working_Memory 0.0515
## Piatoikina et al. 2021 Verbal_Memory Working_Memory 0.0302
## Cruz et al. 2021 Executive Global_Cognitive_Score -0.0754
## Cruz et al. 2021 Global_Cognitive_Score Processing_Speed 0.0219
## Cruz et al. 2021 Global_Cognitive_Score Verbal_Memory 0.0432
## Cruz et al. 2021 Executive Processing_Speed -0.0535
## Cruz et al. 2021 Executive Verbal_Memory -0.0323
## Piatoikina et al. 2021 Executive Processing_Speed -0.0535
## Piatoikina et al. 2021 Executive Verbal_Memory -0.0323
## Cruz et al. 2021 Processing_Speed Verbal_Memory 0.0213
## Piatoikina et al. 2021 Processing_Speed Verbal_Memory 0.0213
## 95%-CI
## Raffa et al. 2011 [-0.2291; 0.0928]
## Guidara et al. 2020 [-0.2887; 0.0899]
## Guidara et al. 2020 [-0.2201; 0.1983]
## Guidara et al. 2020 [-0.2291; 0.0928]
## Guidara et al. 2020 [-0.1091; 0.2401]
## Cruz et al. 2021 [-0.3143; 0.1417]
## Cruz et al. 2021 [-0.2201; 0.1983]
## Cruz et al. 2021 [-0.2291; 0.0928]
## Cruz et al. 2021 [-0.1953; 0.2608]
## Cruz et al. 2021 [-0.1740; 0.2821]
## Cruz et al. 2021 [-0.1437; 0.3123]
## Piatoikina et al. 2023 [-0.2887; 0.0899]
## Piatoikina et al. 2023 [-0.2291; 0.0928]
## Piatoikina et al. 2023 [-0.1091; 0.2401]
## Altuntas et al. 2000 [-0.1091; 0.2401]
## Guidara et al. 2020 [-0.2238; 0.1613]
## Guidara et al. 2020 [-0.1534; 0.2680]
## Guidara et al. 2020 [-0.1882; 0.1829]
## Cruz et al. 2021 [-0.2467; 0.2103]
## Cruz et al. 2021 [-0.1534; 0.2680]
## Cruz et al. 2021 [-0.2638; 0.1931]
## Cruz et al. 2021 [-0.2426; 0.2144]
## Cruz et al. 2021 [-0.2123; 0.2447]
## Piatoikina et al. 2023 [-0.2238; 0.1613]
## Piatoikina et al. 2023 [-0.1882; 0.1829]
## Guidara et al. 2020 [-0.3302; 0.1532]
## Guidara et al. 2020 [-0.2325; 0.1647]
## Piatoikina et al. 2023 [-0.2325; 0.1647]
## Guidara et al. 2020 [-0.1807; 0.2899]
## Cruz et al. 2021 [-0.1961; 0.1921]
## Cruz et al. 2021 [-0.1678; 0.3146]
## Cruz et al. 2021 [-0.1426; 0.2456]
## Cruz et al. 2021 [-0.1638; 0.2243]
## Piatoikina et al. 2021 [-0.1961; 0.1921]
## Piatoikina et al. 2021 [-0.1426; 0.2456]
## Piatoikina et al. 2021 [-0.1638; 0.2243]
## Cruz et al. 2021 [-0.3166; 0.1657]
## Cruz et al. 2021 [-0.2193; 0.2631]
## Cruz et al. 2021 [-0.1980; 0.2843]
## Cruz et al. 2021 [-0.2476; 0.1405]
## Cruz et al. 2021 [-0.2263; 0.1618]
## Piatoikina et al. 2021 [-0.2476; 0.1405]
## Piatoikina et al. 2021 [-0.2263; 0.1618]
## Cruz et al. 2021 [-0.1728; 0.2153]
## Piatoikina et al. 2021 [-0.1728; 0.2153]
##
## Number of studies: k = 6
## Number of pairwise comparisons: m = 45
## Number of treatments: n = 9
## Number of designs: d = 6
##
## Random effects model
##
## Treatment estimate (sm = 'SMD', comparison: other treatments vs 'Executive'):
## SMD 95%-CI z p-value
## Executive . . . .
## General -0.0131 [-0.2824; 0.2563] -0.10 0.9242
## Global_Cognitive_Score 0.0754 [-0.1657; 0.3166] 0.61 0.5399
## Negative 0.0182 [-0.2103; 0.2467] 0.16 0.8762
## Positive 0.0863 [-0.1417; 0.3143] 0.74 0.4582
## Processing_Speed 0.0535 [-0.1405; 0.2476] 0.54 0.5888
## Total 0.0208 [-0.2416; 0.2832] 0.16 0.8765
## Verbal_Memory 0.0323 [-0.1618; 0.2263] 0.33 0.7446
## Working_Memory 0.0020 [-0.1921; 0.1961] 0.02 0.9838
##
## Quantifying heterogeneity / inconsistency:
## tau^2 = 0; tau = 0; I^2 = 0% [0.0%; 60.2%]
##
## Tests of heterogeneity (within designs) and inconsistency (between designs):
## Q d.f. p-value
## Total 9.6 10 0.4765
## Within designs 0.0 0 --
## Between designs 9.6 10 0.4765
##
## Details of network meta-analysis methods:
## - Frequentist graph-theoretical approach
## - DerSimonian-Laird estimator for tau^2
## - Calculation of I^2 based on Q
##
## === Network Meta-Analysis for Measurement: GSHt ===
## Original data (with adjusted standard errors for multi-arm studies):
##
## treat1 treat2
## Raffa et al. 2011 Negative Positive
## Tsai et al. 2013 Negative Positive
## Tsai et al. 2013 Positive Total
## Nucifora et al. 2017 General Positive
## Nucifora et al. 2017 Global_Cognitive_Score Positive
## Nucifora et al. 2017 Negative Positive
## Nucifora et al. 2017 Positive Total
## Hendouei et al. 2018 General Positive
## Hendouei et al. 2018 Negative Positive
## Hendouei et al. 2018 Positive Total
## Hendouei et al. 2018* General Positive
## Hendouei et al. 2018* Negative Positive
## Hendouei et al. 2018* Positive Total
## Hendouei et al. 2018** General Positive
## Hendouei et al. 2018** Negative Positive
## Hendouei et al. 2018** Positive Total
## Chien et al. 2020 Negative Positive
## Chien et al. 2020* Negative Positive
## Coughlin et al. 2021 Cognitive_Flexibility Positive
## Coughlin et al. 2021 Executive Positive
## Coughlin et al. 2021 Global_Cognitive_Score Positive
## Coughlin et al. 2021 Negative Positive
## Coughlin et al. 2021 Positive Processing_Speed
## Gares-Caballer et al. 2022 CGI Positive
## Gares-Caballer et al. 2022 Executive Positive
## Gares-Caballer et al. 2022 General Positive
## Gares-Caballer et al. 2022 Global_Cognitive_Score Positive
## Gares-Caballer et al. 2022 Negative Positive
## Gares-Caballer et al. 2022 Positive Processing_Speed
## Lin et al. 2023 (N=92) Cognitive_Flexibility Positive
## Lin et al. 2023 (N=92) Functioning Positive
## Lin et al. 2023 (N=92) General Positive
## Lin et al. 2023 (N=92) Negative Positive
## Lin et al. 2023 (N=92) Positive Processing_Speed
## Lin et al. 2023 (N=92) Positive Total
## Lin et al. 2023 (N=92) Positive Working_Memory
## Lin et al. 2023 (N=219) Cognitive_Flexibility Positive
## Lin et al. 2023 (N=219) Functioning Positive
## Lin et al. 2023 (N=219) General Positive
## Lin et al. 2023 (N=219) Negative Positive
## Lin et al. 2023 (N=219) Positive Processing_Speed
## Lin et al. 2023 (N=219) Positive Total
## Lin et al. 2023 (N=219) Positive Working_Memory
## Fathy et al. 2015 Negative Positive
## Kizilpinar et al. 2023 General Positive
## Kizilpinar et al. 2023 Negative Positive
## Kizilpinar et al. 2023 Positive Total
## Tsai et al. 2013 Negative Total
## Nucifora et al. 2017 General Negative
## Nucifora et al. 2017 Global_Cognitive_Score Negative
## Nucifora et al. 2017 Negative Total
## Hendouei et al. 2018 General Negative
## Hendouei et al. 2018 Negative Total
## Hendouei et al. 2018* General Negative
## Hendouei et al. 2018* Negative Total
## Hendouei et al. 2018** General Negative
## Hendouei et al. 2018** Negative Total
## Coughlin et al. 2021 Cognitive_Flexibility Negative
## Coughlin et al. 2021 Executive Negative
## Coughlin et al. 2021 Global_Cognitive_Score Negative
## Coughlin et al. 2021 Negative Processing_Speed
## Gares-Caballer et al. 2022 CGI Negative
## Gares-Caballer et al. 2022 Executive Negative
## Gares-Caballer et al. 2022 General Negative
## Gares-Caballer et al. 2022 Global_Cognitive_Score Negative
## Gares-Caballer et al. 2022 Negative Processing_Speed
## Lin et al. 2023 (N=92) Cognitive_Flexibility Negative
## Lin et al. 2023 (N=92) Functioning Negative
## Lin et al. 2023 (N=92) General Negative
## Lin et al. 2023 (N=92) Negative Processing_Speed
## Lin et al. 2023 (N=92) Negative Total
## Lin et al. 2023 (N=92) Negative Working_Memory
## Lin et al. 2023 (N=219) Cognitive_Flexibility Negative
## Lin et al. 2023 (N=219) Functioning Negative
## Lin et al. 2023 (N=219) General Negative
## Lin et al. 2023 (N=219) Negative Processing_Speed
## Lin et al. 2023 (N=219) Negative Total
## Lin et al. 2023 (N=219) Negative Working_Memory
## Kizilpinar et al. 2023 General Negative
## Kizilpinar et al. 2023 Negative Total
## Nucifora et al. 2017 General Global_Cognitive_Score
## Nucifora et al. 2017 General Total
## Hendouei et al. 2018 General Total
## Hendouei et al. 2018* General Total
## Hendouei et al. 2018** General Total
## Gares-Caballer et al. 2022 CGI General
## Gares-Caballer et al. 2022 Executive General
## Gares-Caballer et al. 2022 General Global_Cognitive_Score
## Gares-Caballer et al. 2022 General Processing_Speed
## Lin et al. 2023 (N=92) Cognitive_Flexibility General
## Lin et al. 2023 (N=92) Functioning General
## Lin et al. 2023 (N=92) General Processing_Speed
## Lin et al. 2023 (N=92) General Total
## Lin et al. 2023 (N=92) General Working_Memory
## Lin et al. 2023 (N=219) Cognitive_Flexibility General
## Lin et al. 2023 (N=219) Functioning General
## Lin et al. 2023 (N=219) General Processing_Speed
## Lin et al. 2023 (N=219) General Total
## Lin et al. 2023 (N=219) General Working_Memory
## Kizilpinar et al. 2023 General Total
## Nucifora et al. 2017 Global_Cognitive_Score Total
## Lin et al. 2023 (N=92) Cognitive_Flexibility Total
## Lin et al. 2023 (N=92) Functioning Total
## Lin et al. 2023 (N=92) Processing_Speed Total
## Lin et al. 2023 (N=92) Total Working_Memory
## Lin et al. 2023 (N=219) Cognitive_Flexibility Total
## Lin et al. 2023 (N=219) Functioning Total
## Lin et al. 2023 (N=219) Processing_Speed Total
## Lin et al. 2023 (N=219) Total Working_Memory
## Gonzalez-Liencres et al. 2014 Cognitive_Flexibility Executive
## Gonzalez-Liencres et al. 2014 Executive Processing_Speed
## Coughlin et al. 2021 Cognitive_Flexibility Executive
## Coughlin et al. 2021 Executive Global_Cognitive_Score
## Coughlin et al. 2021 Executive Processing_Speed
## Gares-Caballer et al. 2022 CGI Executive
## Gares-Caballer et al. 2022 Executive Global_Cognitive_Score
## Gares-Caballer et al. 2022 Executive Processing_Speed
## Coughlin et al. 2021 Cognitive_Flexibility Global_Cognitive_Score
## Coughlin et al. 2021 Global_Cognitive_Score Processing_Speed
## Gares-Caballer et al. 2022 CGI Global_Cognitive_Score
## Gares-Caballer et al. 2022 Global_Cognitive_Score Processing_Speed
## Gonzalez-Liencres et al. 2014 Cognitive_Flexibility Processing_Speed
## Coughlin et al. 2021 Cognitive_Flexibility Processing_Speed
## Gares-Caballer et al. 2022 CGI Processing_Speed
## Lin et al. 2023 (N=92) Cognitive_Flexibility Processing_Speed
## Lin et al. 2023 (N=92) Functioning Processing_Speed
## Lin et al. 2023 (N=92) Processing_Speed Working_Memory
## Lin et al. 2023 (N=219) Cognitive_Flexibility Processing_Speed
## Lin et al. 2023 (N=219) Functioning Processing_Speed
## Lin et al. 2023 (N=219) Processing_Speed Working_Memory
## Lin et al. 2023 (N=92) Cognitive_Flexibility Functioning
## Lin et al. 2023 (N=92) Cognitive_Flexibility Working_Memory
## Lin et al. 2023 (N=219) Cognitive_Flexibility Functioning
## Lin et al. 2023 (N=219) Cognitive_Flexibility Working_Memory
## Lin et al. 2023 (N=92) Functioning Working_Memory
## Lin et al. 2023 (N=219) Functioning Working_Memory
## TE seTE seTE.adj narms multiarm
## Raffa et al. 2011 -0.5693 0.3162 0.3396 2
## Tsai et al. 2013 -0.0504 0.2294 0.3193 3 *
## Tsai et al. 2013 0.1253 0.2294 0.3193 3 *
## Nucifora et al. 2017 0.1075 0.2041 0.3775 5 *
## Nucifora et al. 2017 0.6258 0.2041 0.3775 5 *
## Nucifora et al. 2017 0.1699 0.2041 0.3775 5 *
## Nucifora et al. 2017 -0.0541 0.2041 0.3775 5 *
## Hendouei et al. 2018 -0.0000 0.2540 0.3996 4 *
## Hendouei et al. 2018 0.1705 0.2540 0.3996 4 *
## Hendouei et al. 2018 0.0000 0.2540 0.3996 4 *
## Hendouei et al. 2018* 0.1024 0.2540 0.3996 4 *
## Hendouei et al. 2018* 0.1024 0.2540 0.3996 4 *
## Hendouei et al. 2018* -0.1024 0.2540 0.3996 4 *
## Hendouei et al. 2018** -0.0000 0.2626 0.4106 4 *
## Hendouei et al. 2018** -0.3031 0.2626 0.4106 4 *
## Hendouei et al. 2018** 0.1024 0.2626 0.4106 4 *
## Chien et al. 2020 -0.0300 0.2236 0.2556 2
## Chien et al. 2020* -0.3953 0.3536 0.3746 2
## Coughlin et al. 2021 0.1531 0.3086 0.5759 6 *
## Coughlin et al. 2021 0.6979 0.3086 0.5759 6 *
## Coughlin et al. 2021 0.8607 0.3086 0.5759 6 *
## Coughlin et al. 2021 0.1531 0.3086 0.5759 6 *
## Coughlin et al. 2021 -0.6488 0.3086 0.5759 6 *
## Gares-Caballer et al. 2022 -0.0201 0.2722 0.5594 7 *
## Gares-Caballer et al. 2022 0.4837 0.2722 0.5594 7 *
## Gares-Caballer et al. 2022 0.0611 0.2722 0.5594 7 *
## Gares-Caballer et al. 2022 0.4142 0.2722 0.5594 7 *
## Gares-Caballer et al. 2022 0.0701 0.2722 0.5594 7 *
## Gares-Caballer et al. 2022 -0.2732 0.2722 0.5594 7 *
## Lin et al. 2023 (N=92) -0.1372 0.1499 0.3889 8 *
## Lin et al. 2023 (N=92) -0.1856 0.1499 0.3889 8 *
## Lin et al. 2023 (N=92) 0.0679 0.1499 0.3889 8 *
## Lin et al. 2023 (N=92) -0.1052 0.1499 0.3889 8 *
## Lin et al. 2023 (N=92) 0.2519 0.1499 0.3889 8 *
## Lin et al. 2023 (N=92) 0.0101 0.1499 0.3889 8 *
## Lin et al. 2023 (N=92) 0.1724 0.1499 0.3889 8 *
## Lin et al. 2023 (N=219) -0.0171 0.0962 0.3136 8 *
## Lin et al. 2023 (N=219) 0.0050 0.0962 0.3136 8 *
## Lin et al. 2023 (N=219) -0.1704 0.0962 0.3136 8 *
## Lin et al. 2023 (N=219) -0.1262 0.0962 0.3136 8 *
## Lin et al. 2023 (N=219) 0.0261 0.0962 0.3136 8 *
## Lin et al. 2023 (N=219) 0.1182 0.0962 0.3136 8 *
## Lin et al. 2023 (N=219) 0.0981 0.0962 0.3136 8 *
## Fathy et al. 2015 -1.4399 0.2722 0.2990 2
## Kizilpinar et al. 2023 0.1769 0.2949 0.4523 4 *
## Kizilpinar et al. 2023 0.1019 0.2949 0.4523 4 *
## Kizilpinar et al. 2023 -0.1559 0.2949 0.4523 4 *
## Tsai et al. 2013 0.0749 0.2294 0.3193 3 *
## Nucifora et al. 2017 -0.0624 0.2041 0.3775 5 *
## Nucifora et al. 2017 0.4559 0.2041 0.3775 5 *
## Nucifora et al. 2017 0.1158 0.2041 0.3775 5 *
## Hendouei et al. 2018 -0.1705 0.2540 0.3996 4 *
## Hendouei et al. 2018 0.1705 0.2540 0.3996 4 *
## Hendouei et al. 2018* -0.0000 0.2540 0.3996 4 *
## Hendouei et al. 2018* 0.0000 0.2540 0.3996 4 *
## Hendouei et al. 2018** 0.3031 0.2626 0.4106 4 *
## Hendouei et al. 2018** -0.2007 0.2626 0.4106 4 *
## Coughlin et al. 2021 -0.0000 0.3086 0.5759 6 *
## Coughlin et al. 2021 0.5448 0.3086 0.5759 6 *
## Coughlin et al. 2021 0.7076 0.3086 0.5759 6 *
## Coughlin et al. 2021 -0.4957 0.3086 0.5759 6 *
## Gares-Caballer et al. 2022 -0.0902 0.2722 0.5594 7 *
## Gares-Caballer et al. 2022 0.4136 0.2722 0.5594 7 *
## Gares-Caballer et al. 2022 -0.0090 0.2722 0.5594 7 *
## Gares-Caballer et al. 2022 0.3441 0.2722 0.5594 7 *
## Gares-Caballer et al. 2022 -0.2032 0.2722 0.5594 7 *
## Lin et al. 2023 (N=92) -0.0321 0.1499 0.3889 8 *
## Lin et al. 2023 (N=92) -0.0804 0.1499 0.3889 8 *
## Lin et al. 2023 (N=92) 0.1730 0.1499 0.3889 8 *
## Lin et al. 2023 (N=92) 0.1467 0.1499 0.3889 8 *
## Lin et al. 2023 (N=92) -0.0951 0.1499 0.3889 8 *
## Lin et al. 2023 (N=92) 0.0673 0.1499 0.3889 8 *
## Lin et al. 2023 (N=219) 0.1091 0.0962 0.3136 8 *
## Lin et al. 2023 (N=219) 0.1312 0.0962 0.3136 8 *
## Lin et al. 2023 (N=219) -0.0443 0.0962 0.3136 8 *
## Lin et al. 2023 (N=219) -0.1001 0.0962 0.3136 8 *
## Lin et al. 2023 (N=219) -0.0080 0.0962 0.3136 8 *
## Lin et al. 2023 (N=219) -0.0280 0.0962 0.3136 8 *
## Kizilpinar et al. 2023 0.0750 0.2949 0.4523 4 *
## Kizilpinar et al. 2023 -0.0540 0.2949 0.4523 4 *
## Nucifora et al. 2017 -0.5183 0.2041 0.3775 5 *
## Nucifora et al. 2017 0.0534 0.2041 0.3775 5 *
## Hendouei et al. 2018 0.0000 0.2540 0.3996 4 *
## Hendouei et al. 2018* 0.0000 0.2540 0.3996 4 *
## Hendouei et al. 2018** 0.1024 0.2626 0.4106 4 *
## Gares-Caballer et al. 2022 -0.0812 0.2722 0.5594 7 *
## Gares-Caballer et al. 2022 0.4226 0.2722 0.5594 7 *
## Gares-Caballer et al. 2022 -0.3531 0.2722 0.5594 7 *
## Gares-Caballer et al. 2022 -0.2122 0.2722 0.5594 7 *
## Lin et al. 2023 (N=92) -0.2051 0.1499 0.3889 8 *
## Lin et al. 2023 (N=92) -0.2534 0.1499 0.3889 8 *
## Lin et al. 2023 (N=92) 0.3198 0.1499 0.3889 8 *
## Lin et al. 2023 (N=92) 0.0779 0.1499 0.3889 8 *
## Lin et al. 2023 (N=92) 0.2403 0.1499 0.3889 8 *
## Lin et al. 2023 (N=219) 0.1534 0.0962 0.3136 8 *
## Lin et al. 2023 (N=219) 0.1755 0.0962 0.3136 8 *
## Lin et al. 2023 (N=219) -0.1444 0.0962 0.3136 8 *
## Lin et al. 2023 (N=219) -0.0523 0.0962 0.3136 8 *
## Lin et al. 2023 (N=219) -0.0723 0.0962 0.3136 8 *
## Kizilpinar et al. 2023 0.0210 0.2949 0.4523 4 *
## Nucifora et al. 2017 0.5717 0.2041 0.3775 5 *
## Lin et al. 2023 (N=92) -0.1272 0.1499 0.3889 8 *
## Lin et al. 2023 (N=92) -0.1755 0.1499 0.3889 8 *
## Lin et al. 2023 (N=92) -0.2418 0.1499 0.3889 8 *
## Lin et al. 2023 (N=92) 0.1624 0.1499 0.3889 8 *
## Lin et al. 2023 (N=219) 0.1011 0.0962 0.3136 8 *
## Lin et al. 2023 (N=219) 0.1232 0.0962 0.3136 8 *
## Lin et al. 2023 (N=219) 0.0921 0.0962 0.3136 8 *
## Lin et al. 2023 (N=219) -0.0200 0.0962 0.3136 8 *
## Gonzalez-Liencres et al. 2014 0.0300 0.2294 0.3193 3 *
## Gonzalez-Liencres et al. 2014 0.2923 0.2294 0.3193 3 *
## Coughlin et al. 2021 -0.5448 0.3086 0.5759 6 *
## Coughlin et al. 2021 -0.1628 0.3086 0.5759 6 *
## Coughlin et al. 2021 0.0491 0.3086 0.5759 6 *
## Gares-Caballer et al. 2022 -0.5038 0.2722 0.5594 7 *
## Gares-Caballer et al. 2022 0.0696 0.2722 0.5594 7 *
## Gares-Caballer et al. 2022 0.2105 0.2722 0.5594 7 *
## Coughlin et al. 2021 -0.7076 0.3086 0.5759 6 *
## Coughlin et al. 2021 0.2119 0.3086 0.5759 6 *
## Gares-Caballer et al. 2022 -0.4343 0.2722 0.5594 7 *
## Gares-Caballer et al. 2022 0.1409 0.2722 0.5594 7 *
## Gonzalez-Liencres et al. 2014 0.3223 0.2294 0.3193 3 *
## Coughlin et al. 2021 -0.4957 0.3086 0.5759 6 *
## Gares-Caballer et al. 2022 -0.2933 0.2722 0.5594 7 *
## Lin et al. 2023 (N=92) 0.1147 0.1499 0.3889 8 *
## Lin et al. 2023 (N=92) 0.0663 0.1499 0.3889 8 *
## Lin et al. 2023 (N=92) -0.0794 0.1499 0.3889 8 *
## Lin et al. 2023 (N=219) 0.0090 0.0962 0.3136 8 *
## Lin et al. 2023 (N=219) 0.0311 0.0962 0.3136 8 *
## Lin et al. 2023 (N=219) 0.0720 0.0962 0.3136 8 *
## Lin et al. 2023 (N=92) 0.0483 0.1499 0.3889 8 *
## Lin et al. 2023 (N=92) 0.0352 0.1499 0.3889 8 *
## Lin et al. 2023 (N=219) -0.0221 0.0962 0.3136 8 *
## Lin et al. 2023 (N=219) 0.0811 0.0962 0.3136 8 *
## Lin et al. 2023 (N=92) -0.0131 0.1499 0.3889 8 *
## Lin et al. 2023 (N=219) 0.1032 0.0962 0.3136 8 *
##
## Number of treatment arms (by study):
## narms
## Raffa et al. 2011 2
## Tsai et al. 2013 3
## Chien et al. 2020 2
## Chien et al. 2020* 2
## Fathy et al. 2015 2
## Hendouei et al. 2018 4
## Hendouei et al. 2018* 4
## Hendouei et al. 2018** 4
## Kizilpinar et al. 2023 4
## Nucifora et al. 2017 5
## Gonzalez-Liencres et al. 2014 3
## Coughlin et al. 2021 6
## Gares-Caballer et al. 2022 7
## Lin et al. 2023 (N=92) 8
## Lin et al. 2023 (N=219) 8
##
## Results (random effects model):
##
## treat1 treat2
## Raffa et al. 2011 Negative Positive
## Tsai et al. 2013 Negative Positive
## Tsai et al. 2013 Positive Total
## Nucifora et al. 2017 General Positive
## Nucifora et al. 2017 Global_Cognitive_Score Positive
## Nucifora et al. 2017 Negative Positive
## Nucifora et al. 2017 Positive Total
## Hendouei et al. 2018 General Positive
## Hendouei et al. 2018 Negative Positive
## Hendouei et al. 2018 Positive Total
## Hendouei et al. 2018* General Positive
## Hendouei et al. 2018* Negative Positive
## Hendouei et al. 2018* Positive Total
## Hendouei et al. 2018** General Positive
## Hendouei et al. 2018** Negative Positive
## Hendouei et al. 2018** Positive Total
## Chien et al. 2020 Negative Positive
## Chien et al. 2020* Negative Positive
## Coughlin et al. 2021 Cognitive_Flexibility Positive
## Coughlin et al. 2021 Executive Positive
## Coughlin et al. 2021 Global_Cognitive_Score Positive
## Coughlin et al. 2021 Negative Positive
## Coughlin et al. 2021 Positive Processing_Speed
## Gares-Caballer et al. 2022 CGI Positive
## Gares-Caballer et al. 2022 Executive Positive
## Gares-Caballer et al. 2022 General Positive
## Gares-Caballer et al. 2022 Global_Cognitive_Score Positive
## Gares-Caballer et al. 2022 Negative Positive
## Gares-Caballer et al. 2022 Positive Processing_Speed
## Lin et al. 2023 (N=92) Cognitive_Flexibility Positive
## Lin et al. 2023 (N=92) Functioning Positive
## Lin et al. 2023 (N=92) General Positive
## Lin et al. 2023 (N=92) Negative Positive
## Lin et al. 2023 (N=92) Positive Processing_Speed
## Lin et al. 2023 (N=92) Positive Total
## Lin et al. 2023 (N=92) Positive Working_Memory
## Lin et al. 2023 (N=219) Cognitive_Flexibility Positive
## Lin et al. 2023 (N=219) Functioning Positive
## Lin et al. 2023 (N=219) General Positive
## Lin et al. 2023 (N=219) Negative Positive
## Lin et al. 2023 (N=219) Positive Processing_Speed
## Lin et al. 2023 (N=219) Positive Total
## Lin et al. 2023 (N=219) Positive Working_Memory
## Fathy et al. 2015 Negative Positive
## Kizilpinar et al. 2023 General Positive
## Kizilpinar et al. 2023 Negative Positive
## Kizilpinar et al. 2023 Positive Total
## Tsai et al. 2013 Negative Total
## Nucifora et al. 2017 General Negative
## Nucifora et al. 2017 Global_Cognitive_Score Negative
## Nucifora et al. 2017 Negative Total
## Hendouei et al. 2018 General Negative
## Hendouei et al. 2018 Negative Total
## Hendouei et al. 2018* General Negative
## Hendouei et al. 2018* Negative Total
## Hendouei et al. 2018** General Negative
## Hendouei et al. 2018** Negative Total
## Coughlin et al. 2021 Cognitive_Flexibility Negative
## Coughlin et al. 2021 Executive Negative
## Coughlin et al. 2021 Global_Cognitive_Score Negative
## Coughlin et al. 2021 Negative Processing_Speed
## Gares-Caballer et al. 2022 CGI Negative
## Gares-Caballer et al. 2022 Executive Negative
## Gares-Caballer et al. 2022 General Negative
## Gares-Caballer et al. 2022 Global_Cognitive_Score Negative
## Gares-Caballer et al. 2022 Negative Processing_Speed
## Lin et al. 2023 (N=92) Cognitive_Flexibility Negative
## Lin et al. 2023 (N=92) Functioning Negative
## Lin et al. 2023 (N=92) General Negative
## Lin et al. 2023 (N=92) Negative Processing_Speed
## Lin et al. 2023 (N=92) Negative Total
## Lin et al. 2023 (N=92) Negative Working_Memory
## Lin et al. 2023 (N=219) Cognitive_Flexibility Negative
## Lin et al. 2023 (N=219) Functioning Negative
## Lin et al. 2023 (N=219) General Negative
## Lin et al. 2023 (N=219) Negative Processing_Speed
## Lin et al. 2023 (N=219) Negative Total
## Lin et al. 2023 (N=219) Negative Working_Memory
## Kizilpinar et al. 2023 General Negative
## Kizilpinar et al. 2023 Negative Total
## Nucifora et al. 2017 General Global_Cognitive_Score
## Nucifora et al. 2017 General Total
## Hendouei et al. 2018 General Total
## Hendouei et al. 2018* General Total
## Hendouei et al. 2018** General Total
## Gares-Caballer et al. 2022 CGI General
## Gares-Caballer et al. 2022 Executive General
## Gares-Caballer et al. 2022 General Global_Cognitive_Score
## Gares-Caballer et al. 2022 General Processing_Speed
## Lin et al. 2023 (N=92) Cognitive_Flexibility General
## Lin et al. 2023 (N=92) Functioning General
## Lin et al. 2023 (N=92) General Processing_Speed
## Lin et al. 2023 (N=92) General Total
## Lin et al. 2023 (N=92) General Working_Memory
## Lin et al. 2023 (N=219) Cognitive_Flexibility General
## Lin et al. 2023 (N=219) Functioning General
## Lin et al. 2023 (N=219) General Processing_Speed
## Lin et al. 2023 (N=219) General Total
## Lin et al. 2023 (N=219) General Working_Memory
## Kizilpinar et al. 2023 General Total
## Nucifora et al. 2017 Global_Cognitive_Score Total
## Lin et al. 2023 (N=92) Cognitive_Flexibility Total
## Lin et al. 2023 (N=92) Functioning Total
## Lin et al. 2023 (N=92) Processing_Speed Total
## Lin et al. 2023 (N=92) Total Working_Memory
## Lin et al. 2023 (N=219) Cognitive_Flexibility Total
## Lin et al. 2023 (N=219) Functioning Total
## Lin et al. 2023 (N=219) Processing_Speed Total
## Lin et al. 2023 (N=219) Total Working_Memory
## Gonzalez-Liencres et al. 2014 Cognitive_Flexibility Executive
## Gonzalez-Liencres et al. 2014 Executive Processing_Speed
## Coughlin et al. 2021 Cognitive_Flexibility Executive
## Coughlin et al. 2021 Executive Global_Cognitive_Score
## Coughlin et al. 2021 Executive Processing_Speed
## Gares-Caballer et al. 2022 CGI Executive
## Gares-Caballer et al. 2022 Executive Global_Cognitive_Score
## Gares-Caballer et al. 2022 Executive Processing_Speed
## Coughlin et al. 2021 Cognitive_Flexibility Global_Cognitive_Score
## Coughlin et al. 2021 Global_Cognitive_Score Processing_Speed
## Gares-Caballer et al. 2022 CGI Global_Cognitive_Score
## Gares-Caballer et al. 2022 Global_Cognitive_Score Processing_Speed
## Gonzalez-Liencres et al. 2014 Cognitive_Flexibility Processing_Speed
## Coughlin et al. 2021 Cognitive_Flexibility Processing_Speed
## Gares-Caballer et al. 2022 CGI Processing_Speed
## Lin et al. 2023 (N=92) Cognitive_Flexibility Processing_Speed
## Lin et al. 2023 (N=92) Functioning Processing_Speed
## Lin et al. 2023 (N=92) Processing_Speed Working_Memory
## Lin et al. 2023 (N=219) Cognitive_Flexibility Processing_Speed
## Lin et al. 2023 (N=219) Functioning Processing_Speed
## Lin et al. 2023 (N=219) Processing_Speed Working_Memory
## Lin et al. 2023 (N=92) Cognitive_Flexibility Functioning
## Lin et al. 2023 (N=92) Cognitive_Flexibility Working_Memory
## Lin et al. 2023 (N=219) Cognitive_Flexibility Functioning
## Lin et al. 2023 (N=219) Cognitive_Flexibility Working_Memory
## Lin et al. 2023 (N=92) Functioning Working_Memory
## Lin et al. 2023 (N=219) Functioning Working_Memory
## SMD 95%-CI
## Raffa et al. 2011 -0.1295 [-0.2652; 0.0062]
## Tsai et al. 2013 -0.1295 [-0.2652; 0.0062]
## Tsai et al. 2013 0.0638 [-0.0907; 0.2184]
## Nucifora et al. 2017 -0.0414 [-0.1975; 0.1147]
## Nucifora et al. 2017 0.4302 [ 0.1646; 0.6959]
## Nucifora et al. 2017 -0.1295 [-0.2652; 0.0062]
## Nucifora et al. 2017 0.0638 [-0.0907; 0.2184]
## Hendouei et al. 2018 -0.0414 [-0.1975; 0.1147]
## Hendouei et al. 2018 -0.1295 [-0.2652; 0.0062]
## Hendouei et al. 2018 0.0638 [-0.0907; 0.2184]
## Hendouei et al. 2018* -0.0414 [-0.1975; 0.1147]
## Hendouei et al. 2018* -0.1295 [-0.2652; 0.0062]
## Hendouei et al. 2018* 0.0638 [-0.0907; 0.2184]
## Hendouei et al. 2018** -0.0414 [-0.1975; 0.1147]
## Hendouei et al. 2018** -0.1295 [-0.2652; 0.0062]
## Hendouei et al. 2018** 0.0638 [-0.0907; 0.2184]
## Chien et al. 2020 -0.1295 [-0.2652; 0.0062]
## Chien et al. 2020* -0.1295 [-0.2652; 0.0062]
## Coughlin et al. 2021 -0.0127 [-0.2026; 0.1772]
## Coughlin et al. 2021 0.2769 [-0.0150; 0.5688]
## Coughlin et al. 2021 0.4302 [ 0.1646; 0.6959]
## Coughlin et al. 2021 -0.1295 [-0.2652; 0.0062]
## Coughlin et al. 2021 0.0318 [-0.1502; 0.2137]
## Gares-Caballer et al. 2022 -0.1531 [-0.6146; 0.3085]
## Gares-Caballer et al. 2022 0.2769 [-0.0150; 0.5688]
## Gares-Caballer et al. 2022 -0.0414 [-0.1975; 0.1147]
## Gares-Caballer et al. 2022 0.4302 [ 0.1646; 0.6959]
## Gares-Caballer et al. 2022 -0.1295 [-0.2652; 0.0062]
## Gares-Caballer et al. 2022 0.0318 [-0.1502; 0.2137]
## Lin et al. 2023 (N=92) -0.0127 [-0.2026; 0.1772]
## Lin et al. 2023 (N=92) -0.0417 [-0.2488; 0.1654]
## Lin et al. 2023 (N=92) -0.0414 [-0.1975; 0.1147]
## Lin et al. 2023 (N=92) -0.1295 [-0.2652; 0.0062]
## Lin et al. 2023 (N=92) 0.0318 [-0.1502; 0.2137]
## Lin et al. 2023 (N=92) 0.0638 [-0.0907; 0.2184]
## Lin et al. 2023 (N=92) 0.0990 [-0.1080; 0.3061]
## Lin et al. 2023 (N=219) -0.0127 [-0.2026; 0.1772]
## Lin et al. 2023 (N=219) -0.0417 [-0.2488; 0.1654]
## Lin et al. 2023 (N=219) -0.0414 [-0.1975; 0.1147]
## Lin et al. 2023 (N=219) -0.1295 [-0.2652; 0.0062]
## Lin et al. 2023 (N=219) 0.0318 [-0.1502; 0.2137]
## Lin et al. 2023 (N=219) 0.0638 [-0.0907; 0.2184]
## Lin et al. 2023 (N=219) 0.0990 [-0.1080; 0.3061]
## Fathy et al. 2015 -0.1295 [-0.2652; 0.0062]
## Kizilpinar et al. 2023 -0.0414 [-0.1975; 0.1147]
## Kizilpinar et al. 2023 -0.1295 [-0.2652; 0.0062]
## Kizilpinar et al. 2023 0.0638 [-0.0907; 0.2184]
## Tsai et al. 2013 -0.0656 [-0.2202; 0.0889]
## Nucifora et al. 2017 0.0880 [-0.0681; 0.2442]
## Nucifora et al. 2017 0.5597 [ 0.2940; 0.8254]
## Nucifora et al. 2017 -0.0656 [-0.2202; 0.0889]
## Hendouei et al. 2018 0.0880 [-0.0681; 0.2442]
## Hendouei et al. 2018 -0.0656 [-0.2202; 0.0889]
## Hendouei et al. 2018* 0.0880 [-0.0681; 0.2442]
## Hendouei et al. 2018* -0.0656 [-0.2202; 0.0889]
## Hendouei et al. 2018** 0.0880 [-0.0681; 0.2442]
## Hendouei et al. 2018** -0.0656 [-0.2202; 0.0889]
## Coughlin et al. 2021 0.1168 [-0.0731; 0.3067]
## Coughlin et al. 2021 0.4064 [ 0.1145; 0.6983]
## Coughlin et al. 2021 0.5597 [ 0.2940; 0.8254]
## Coughlin et al. 2021 -0.0977 [-0.2797; 0.0843]
## Gares-Caballer et al. 2022 -0.0236 [-0.4851; 0.4380]
## Gares-Caballer et al. 2022 0.4064 [ 0.1145; 0.6983]
## Gares-Caballer et al. 2022 0.0880 [-0.0681; 0.2442]
## Gares-Caballer et al. 2022 0.5597 [ 0.2940; 0.8254]
## Gares-Caballer et al. 2022 -0.0977 [-0.2797; 0.0843]
## Lin et al. 2023 (N=92) 0.1168 [-0.0731; 0.3067]
## Lin et al. 2023 (N=92) 0.0878 [-0.1193; 0.2948]
## Lin et al. 2023 (N=92) 0.0880 [-0.0681; 0.2442]
## Lin et al. 2023 (N=92) -0.0977 [-0.2797; 0.0843]
## Lin et al. 2023 (N=92) -0.0656 [-0.2202; 0.0889]
## Lin et al. 2023 (N=92) -0.0304 [-0.2375; 0.1766]
## Lin et al. 2023 (N=219) 0.1168 [-0.0731; 0.3067]
## Lin et al. 2023 (N=219) 0.0878 [-0.1193; 0.2948]
## Lin et al. 2023 (N=219) 0.0880 [-0.0681; 0.2442]
## Lin et al. 2023 (N=219) -0.0977 [-0.2797; 0.0843]
## Lin et al. 2023 (N=219) -0.0656 [-0.2202; 0.0889]
## Lin et al. 2023 (N=219) -0.0304 [-0.2375; 0.1766]
## Kizilpinar et al. 2023 0.0880 [-0.0681; 0.2442]
## Kizilpinar et al. 2023 -0.0656 [-0.2202; 0.0889]
## Nucifora et al. 2017 -0.4717 [-0.7446; -0.1987]
## Nucifora et al. 2017 0.0224 [-0.1428; 0.1876]
## Hendouei et al. 2018 0.0224 [-0.1428; 0.1876]
## Hendouei et al. 2018* 0.0224 [-0.1428; 0.1876]
## Hendouei et al. 2018** 0.0224 [-0.1428; 0.1876]
## Gares-Caballer et al. 2022 -0.1116 [-0.5763; 0.3531]
## Gares-Caballer et al. 2022 0.3183 [ 0.0189; 0.6178]
## Gares-Caballer et al. 2022 -0.4717 [-0.7446; -0.1987]
## Gares-Caballer et al. 2022 -0.0097 [-0.2004; 0.1811]
## Lin et al. 2023 (N=92) 0.0288 [-0.1696; 0.2272]
## Lin et al. 2023 (N=92) -0.0003 [-0.2137; 0.2131]
## Lin et al. 2023 (N=92) -0.0097 [-0.2004; 0.1811]
## Lin et al. 2023 (N=92) 0.0224 [-0.1428; 0.1876]
## Lin et al. 2023 (N=92) 0.0576 [-0.1558; 0.2710]
## Lin et al. 2023 (N=219) 0.0288 [-0.1696; 0.2272]
## Lin et al. 2023 (N=219) -0.0003 [-0.2137; 0.2131]
## Lin et al. 2023 (N=219) -0.0097 [-0.2004; 0.1811]
## Lin et al. 2023 (N=219) 0.0224 [-0.1428; 0.1876]
## Lin et al. 2023 (N=219) 0.0576 [-0.1558; 0.2710]
## Kizilpinar et al. 2023 0.0224 [-0.1428; 0.1876]
## Nucifora et al. 2017 0.4941 [ 0.2179; 0.7702]
## Lin et al. 2023 (N=92) 0.0512 [-0.1479; 0.2502]
## Lin et al. 2023 (N=92) 0.0221 [-0.1913; 0.2355]
## Lin et al. 2023 (N=92) 0.0321 [-0.1608; 0.2249]
## Lin et al. 2023 (N=92) 0.0352 [-0.1782; 0.2486]
## Lin et al. 2023 (N=219) 0.0512 [-0.1479; 0.2502]
## Lin et al. 2023 (N=219) 0.0221 [-0.1913; 0.2355]
## Lin et al. 2023 (N=219) 0.0321 [-0.1608; 0.2249]
## Lin et al. 2023 (N=219) 0.0352 [-0.1782; 0.2486]
## Gonzalez-Liencres et al. 2014 -0.2896 [-0.5873; 0.0081]
## Gonzalez-Liencres et al. 2014 0.3087 [ 0.0185; 0.5989]
## Coughlin et al. 2021 -0.2896 [-0.5873; 0.0081]
## Coughlin et al. 2021 -0.1533 [-0.5031; 0.1965]
## Coughlin et al. 2021 0.3087 [ 0.0185; 0.5989]
## Gares-Caballer et al. 2022 -0.4300 [-0.9322; 0.0722]
## Gares-Caballer et al. 2022 -0.1533 [-0.5031; 0.1965]
## Gares-Caballer et al. 2022 0.3087 [ 0.0185; 0.5989]
## Coughlin et al. 2021 -0.4429 [-0.7367; -0.1491]
## Coughlin et al. 2021 0.4620 [ 0.1768; 0.7471]
## Gares-Caballer et al. 2022 -0.5833 [-1.0776; -0.0890]
## Gares-Caballer et al. 2022 0.4620 [ 0.1768; 0.7471]
## Gonzalez-Liencres et al. 2014 0.0191 [-0.1819; 0.2201]
## Coughlin et al. 2021 0.0191 [-0.1819; 0.2201]
## Gares-Caballer et al. 2022 -0.1213 [-0.5894; 0.3468]
## Lin et al. 2023 (N=92) 0.0191 [-0.1819; 0.2201]
## Lin et al. 2023 (N=92) -0.0099 [-0.2341; 0.2142]
## Lin et al. 2023 (N=92) 0.0673 [-0.1569; 0.2915]
## Lin et al. 2023 (N=219) 0.0191 [-0.1819; 0.2201]
## Lin et al. 2023 (N=219) -0.0099 [-0.2341; 0.2142]
## Lin et al. 2023 (N=219) 0.0673 [-0.1569; 0.2915]
## Lin et al. 2023 (N=92) 0.0290 [-0.1993; 0.2574]
## Lin et al. 2023 (N=92) 0.0864 [-0.1420; 0.3147]
## Lin et al. 2023 (N=219) 0.0290 [-0.1993; 0.2574]
## Lin et al. 2023 (N=219) 0.0864 [-0.1420; 0.3147]
## Lin et al. 2023 (N=92) 0.0573 [-0.1819; 0.2966]
## Lin et al. 2023 (N=219) 0.0573 [-0.1819; 0.2966]
##
## Number of studies: k = 15
## Number of pairwise comparisons: m = 136
## Number of treatments: n = 11
## Number of designs: d = 8
##
## Random effects model
##
## Treatment estimate (sm = 'SMD', comparison: other treatments vs 'CGI'):
## SMD 95%-CI z p-value
## CGI . . . .
## Cognitive_Flexibility 0.1404 [-0.3380; 0.6188] 0.58 0.5652
## Executive 0.4300 [-0.0722; 0.9322] 1.68 0.0933
## Functioning 0.1114 [-0.3773; 0.6000] 0.45 0.6551
## General 0.1116 [-0.3531; 0.5763] 0.47 0.6378
## Global_Cognitive_Score 0.5833 [ 0.0890; 1.0776] 2.31 0.0207
## Negative 0.0236 [-0.4380; 0.4851] 0.10 0.9202
## Positive 0.1531 [-0.3085; 0.6146] 0.65 0.5157
## Processing_Speed 0.1213 [-0.3468; 0.5894] 0.51 0.6115
## Total 0.0892 [-0.3810; 0.5595] 0.37 0.7100
## Working_Memory 0.0540 [-0.4346; 0.5427] 0.22 0.8284
##
## Quantifying heterogeneity / inconsistency:
## tau^2 = 0.0153; tau = 0.1238; I^2 = 29% [0.0%; 52.1%]
##
## Tests of heterogeneity (within designs) and inconsistency (between designs):
## Q d.f. p-value
## Total 54.97 39 0.0464
## Within designs 29.80 19 0.0544
## Between designs 25.16 20 0.1953
##
## Details of network meta-analysis methods:
## - Frequentist graph-theoretical approach
## - DerSimonian-Laird estimator for tau^2
## - Calculation of I^2 based on Q
##
## === Network Meta-Analysis for Measurement: GSSG ===
## Original data:
##
## treat1 treat2 TE seTE
## Raffa et al. 2011 Negative Positive 0.0103 0.3162
## Tao et al. 2020 Negative Positive -0.2018 0.1516
##
## Number of treatment arms (by study):
## narms
## Raffa et al. 2011 2
## Tao et al. 2020 2
##
## Results (random effects model):
##
## treat1 treat2 SMD 95%-CI
## Raffa et al. 2011 Negative Positive -0.1621 [-0.4301; 0.1058]
## Tao et al. 2020 Negative Positive -0.1621 [-0.4301; 0.1058]
##
## Number of studies: k = 2
## Number of pairwise comparisons: m = 2
## Number of treatments: n = 2
## Number of designs: d = 1
##
## Random effects model
##
## Treatment estimate (sm = 'SMD', comparison: 'Positive' vs 'Negative'):
## SMD 95%-CI z p-value
## Negative . . . .
## Positive 0.1621 [-0.1058; 0.4301] 1.19 0.2357
##
## Quantifying heterogeneity:
## tau^2 = 0; tau = 0; I^2 = 0%
##
## Test of heterogeneity:
## Q d.f. p-value
## 0.37 1 0.5454
##
## Details of network meta-analysis methods:
## - Frequentist graph-theoretical approach
## - DerSimonian-Laird estimator for tau^2
## - Calculation of I^2 based on Q
##
## === Network Meta-Analysis for Measurement: Imaging ===
## Original data (with adjusted standard errors for multi-arm studies):
##
## treat1 treat2 TE seTE
## Matsuzawa et al. 2008 Ideational_Fluency Positive 0.6731 0.2985
## Matsuzawa et al. 2008 Negative Positive -0.2333 0.3430
## Matsuzawa et al. 2008 Positive Processing_Speed -0.3190 0.2985
## Matsuzawa et al. 2008 Positive Total -0.0243 0.3430
## Matsuzawa et al. 2008 Positive Verbal_Memory -0.6419 0.2985
## Reyes-Madrigal et al. 2019 General Positive -1.8050 0.5345
## Reyes-Madrigal et al. 2019 Negative Positive -1.5690 0.5345
## Reyes-Madrigal et al. 2019 Positive Total 1.4982 0.5345
## Iwata et al. 2021 General Positive -0.0710 0.1768
## Iwata et al. 2021 Negative Positive 0.2313 0.1768
## Iwata et al. 2021 Positive Total 0.0000 0.1768
## Coughlin et al. 2021 Ideational_Fluency Positive 0.5680 0.3922
## Coughlin et al. 2021 Negative Positive 0.0722 0.3922
## Coughlin et al. 2021 Positive Processing_Speed -0.1252 0.3922
## Coughlin et al. 2021 Positive Verbal_Memory 0.0203 0.3922
## Lesh et al. 2021 Functioning Positive 0.4597 0.2582
## Lesh et al. 2021 Negative Positive 0.2626 0.2582
## Lesh et al. 2021 Positive Total 0.0293 0.2582
## Matsuzawa et al. 2008 Ideational_Fluency Negative 0.9063 0.2985
## Matsuzawa et al. 2008 Negative Processing_Speed -0.5522 0.2985
## Matsuzawa et al. 2008 Negative Total -0.2575 0.3430
## Matsuzawa et al. 2008 Negative Verbal_Memory -0.8751 0.2985
## Reyes-Madrigal et al. 2019 General Negative -0.2360 0.5345
## Reyes-Madrigal et al. 2019 Negative Total -0.0708 0.5345
## Iwata et al. 2021 General Negative -0.3023 0.1768
## Iwata et al. 2021 Negative Total 0.2313 0.1768
## Coughlin et al. 2021 Ideational_Fluency Negative 0.4958 0.3922
## Coughlin et al. 2021 Negative Processing_Speed -0.0529 0.3922
## Coughlin et al. 2021 Negative Verbal_Memory 0.0926 0.3922
## Lesh et al. 2021 Functioning Negative 0.1972 0.2582
## Lesh et al. 2021 Negative Total 0.2918 0.2582
## Ravanfar et al. 2022 Functioning Negative 0.8504 0.4714
## Ravanfar et al. 2022 Negative Total -0.0690 0.4714
## Reyes-Madrigal et al. 2019 General Total -0.3068 0.5345
## Iwata et al. 2021 General Total -0.0710 0.1768
## Matsuzawa et al. 2008 Ideational_Fluency Total 0.6488 0.2985
## Matsuzawa et al. 2008 Processing_Speed Total 0.2947 0.2985
## Matsuzawa et al. 2008 Total Verbal_Memory -0.6176 0.2985
## Lesh et al. 2021 Functioning Total 0.4890 0.2582
## Ravanfar et al. 2022 Functioning Total 0.7814 0.4714
## Matsuzawa et al. 2008 Ideational_Fluency Processing_Speed 0.3541 0.2462
## Matsuzawa et al. 2008 Ideational_Fluency Verbal_Memory 0.0312 0.2462
## Coughlin et al. 2021 Ideational_Fluency Processing_Speed 0.4428 0.3922
## Coughlin et al. 2021 Ideational_Fluency Verbal_Memory 0.5883 0.3922
## Matsuzawa et al. 2008 Processing_Speed Verbal_Memory -0.3229 0.2462
## Coughlin et al. 2021 Processing_Speed Verbal_Memory 0.1455 0.3922
## seTE.adj narms multiarm
## Matsuzawa et al. 2008 0.6628 6 *
## Matsuzawa et al. 2008 0.8074 6 *
## Matsuzawa et al. 2008 0.6628 6 *
## Matsuzawa et al. 2008 0.8074 6 *
## Matsuzawa et al. 2008 0.6628 6 *
## Reyes-Madrigal et al. 2019 0.8283 4 *
## Reyes-Madrigal et al. 2019 0.8283 4 *
## Reyes-Madrigal et al. 2019 0.8283 4 *
## Iwata et al. 2021 0.4209 4 *
## Iwata et al. 2021 0.4209 4 *
## Iwata et al. 2021 0.4209 4 *
## Coughlin et al. 2021 0.7266 5 *
## Coughlin et al. 2021 0.7266 5 *
## Coughlin et al. 2021 0.7266 5 *
## Coughlin et al. 2021 0.7266 5 *
## Lesh et al. 2021 0.4980 4 *
## Lesh et al. 2021 0.4980 4 *
## Lesh et al. 2021 0.4980 4 *
## Matsuzawa et al. 2008 0.6628 6 *
## Matsuzawa et al. 2008 0.6628 6 *
## Matsuzawa et al. 2008 0.8074 6 *
## Matsuzawa et al. 2008 0.6628 6 *
## Reyes-Madrigal et al. 2019 0.8283 4 *
## Reyes-Madrigal et al. 2019 0.8283 4 *
## Iwata et al. 2021 0.4209 4 *
## Iwata et al. 2021 0.4209 4 *
## Coughlin et al. 2021 0.7266 5 *
## Coughlin et al. 2021 0.7266 5 *
## Coughlin et al. 2021 0.7266 5 *
## Lesh et al. 2021 0.4980 4 *
## Lesh et al. 2021 0.4980 4 *
## Ravanfar et al. 2022 0.6476 3 *
## Ravanfar et al. 2022 0.6476 3 *
## Reyes-Madrigal et al. 2019 0.8283 4 *
## Iwata et al. 2021 0.4209 4 *
## Matsuzawa et al. 2008 0.6628 6 *
## Matsuzawa et al. 2008 0.6628 6 *
## Matsuzawa et al. 2008 0.6628 6 *
## Lesh et al. 2021 0.4980 4 *
## Ravanfar et al. 2022 0.6476 3 *
## Matsuzawa et al. 2008 0.5442 6 *
## Matsuzawa et al. 2008 0.5442 6 *
## Coughlin et al. 2021 0.7266 5 *
## Coughlin et al. 2021 0.7266 5 *
## Matsuzawa et al. 2008 0.5442 6 *
## Coughlin et al. 2021 0.7266 5 *
##
## Number of treatment arms (by study):
## narms
## Matsuzawa et al. 2008 6
## Reyes-Madrigal et al. 2019 4
## Iwata et al. 2021 4
## Coughlin et al. 2021 5
## Lesh et al. 2021 4
## Ravanfar et al. 2022 3
##
## Results (random effects model):
##
## treat1 treat2 SMD
## Matsuzawa et al. 2008 Ideational_Fluency Positive 0.6123
## Matsuzawa et al. 2008 Negative Positive -0.0466
## Matsuzawa et al. 2008 Positive Processing_Speed -0.2264
## Matsuzawa et al. 2008 Positive Total 0.1737
## Matsuzawa et al. 2008 Positive Verbal_Memory -0.3814
## Reyes-Madrigal et al. 2019 General Positive -0.3517
## Reyes-Madrigal et al. 2019 Negative Positive -0.0466
## Reyes-Madrigal et al. 2019 Positive Total 0.1737
## Iwata et al. 2021 General Positive -0.3517
## Iwata et al. 2021 Negative Positive -0.0466
## Iwata et al. 2021 Positive Total 0.1737
## Coughlin et al. 2021 Ideational_Fluency Positive 0.6123
## Coughlin et al. 2021 Negative Positive -0.0466
## Coughlin et al. 2021 Positive Processing_Speed -0.2264
## Coughlin et al. 2021 Positive Verbal_Memory -0.3814
## Lesh et al. 2021 Functioning Positive 0.4208
## Lesh et al. 2021 Negative Positive -0.0466
## Lesh et al. 2021 Positive Total 0.1737
## Matsuzawa et al. 2008 Ideational_Fluency Negative 0.6589
## Matsuzawa et al. 2008 Negative Processing_Speed -0.2730
## Matsuzawa et al. 2008 Negative Total 0.1270
## Matsuzawa et al. 2008 Negative Verbal_Memory -0.4281
## Reyes-Madrigal et al. 2019 General Negative -0.3051
## Reyes-Madrigal et al. 2019 Negative Total 0.1270
## Iwata et al. 2021 General Negative -0.3051
## Iwata et al. 2021 Negative Total 0.1270
## Coughlin et al. 2021 Ideational_Fluency Negative 0.6589
## Coughlin et al. 2021 Negative Processing_Speed -0.2730
## Coughlin et al. 2021 Negative Verbal_Memory -0.4281
## Lesh et al. 2021 Functioning Negative 0.4675
## Lesh et al. 2021 Negative Total 0.1270
## Ravanfar et al. 2022 Functioning Negative 0.4675
## Ravanfar et al. 2022 Negative Total 0.1270
## Reyes-Madrigal et al. 2019 General Total -0.1781
## Iwata et al. 2021 General Total -0.1781
## Matsuzawa et al. 2008 Ideational_Fluency Total 0.7859
## Matsuzawa et al. 2008 Processing_Speed Total 0.4000
## Matsuzawa et al. 2008 Total Verbal_Memory -0.5551
## Lesh et al. 2021 Functioning Total 0.5945
## Ravanfar et al. 2022 Functioning Total 0.5945
## Matsuzawa et al. 2008 Ideational_Fluency Processing_Speed 0.3859
## Matsuzawa et al. 2008 Ideational_Fluency Verbal_Memory 0.2308
## Coughlin et al. 2021 Ideational_Fluency Processing_Speed 0.3859
## Coughlin et al. 2021 Ideational_Fluency Verbal_Memory 0.2308
## Matsuzawa et al. 2008 Processing_Speed Verbal_Memory -0.1550
## Coughlin et al. 2021 Processing_Speed Verbal_Memory -0.1550
## 95%-CI
## Matsuzawa et al. 2008 [ 0.1111; 1.1134]
## Matsuzawa et al. 2008 [-0.3840; 0.2907]
## Matsuzawa et al. 2008 [-0.7275; 0.2748]
## Matsuzawa et al. 2008 [-0.1799; 0.5272]
## Matsuzawa et al. 2008 [-0.8826; 0.1197]
## Reyes-Madrigal et al. 2019 [-0.8213; 0.1178]
## Reyes-Madrigal et al. 2019 [-0.3840; 0.2907]
## Reyes-Madrigal et al. 2019 [-0.1799; 0.5272]
## Iwata et al. 2021 [-0.8213; 0.1178]
## Iwata et al. 2021 [-0.3840; 0.2907]
## Iwata et al. 2021 [-0.1799; 0.5272]
## Coughlin et al. 2021 [ 0.1111; 1.1134]
## Coughlin et al. 2021 [-0.3840; 0.2907]
## Coughlin et al. 2021 [-0.7275; 0.2748]
## Coughlin et al. 2021 [-0.8826; 0.1197]
## Lesh et al. 2021 [-0.1084; 0.9501]
## Lesh et al. 2021 [-0.3840; 0.2907]
## Lesh et al. 2021 [-0.1799; 0.5272]
## Matsuzawa et al. 2008 [ 0.1597; 1.1581]
## Matsuzawa et al. 2008 [-0.7722; 0.2262]
## Matsuzawa et al. 2008 [-0.2136; 0.4677]
## Matsuzawa et al. 2008 [-0.9273; 0.0712]
## Reyes-Madrigal et al. 2019 [-0.7714; 0.1613]
## Reyes-Madrigal et al. 2019 [-0.2136; 0.4677]
## Iwata et al. 2021 [-0.7714; 0.1613]
## Iwata et al. 2021 [-0.2136; 0.4677]
## Coughlin et al. 2021 [ 0.1597; 1.1581]
## Coughlin et al. 2021 [-0.7722; 0.2262]
## Coughlin et al. 2021 [-0.9273; 0.0712]
## Lesh et al. 2021 [-0.0430; 0.9780]
## Lesh et al. 2021 [-0.2136; 0.4677]
## Ravanfar et al. 2022 [-0.0430; 0.9780]
## Ravanfar et al. 2022 [-0.2136; 0.4677]
## Reyes-Madrigal et al. 2019 [-0.6484; 0.2923]
## Iwata et al. 2021 [-0.6484; 0.2923]
## Matsuzawa et al. 2008 [ 0.2621; 1.3097]
## Matsuzawa et al. 2008 [-0.1238; 0.9238]
## Matsuzawa et al. 2008 [-1.0789; -0.0313]
## Lesh et al. 2021 [ 0.0814; 1.1076]
## Ravanfar et al. 2022 [ 0.0814; 1.1076]
## Matsuzawa et al. 2008 [-0.1533; 0.9251]
## Matsuzawa et al. 2008 [-0.3083; 0.7700]
## Coughlin et al. 2021 [-0.1533; 0.9251]
## Coughlin et al. 2021 [-0.3083; 0.7700]
## Matsuzawa et al. 2008 [-0.6942; 0.3841]
## Coughlin et al. 2021 [-0.6942; 0.3841]
##
## Number of studies: k = 6
## Number of pairwise comparisons: m = 46
## Number of treatments: n = 8
## Number of designs: d = 5
##
## Random effects model
##
## Treatment estimate (sm = 'SMD', comparison: other treatments vs 'Functioning'):
## SMD 95%-CI z p-value
## Functioning . . . .
## General -0.7726 [-1.4120; -0.1332] -2.37 0.0179
## Ideational_Fluency 0.1914 [-0.4783; 0.8611] 0.56 0.5753
## Negative -0.4675 [-0.9780; 0.0430] -1.79 0.0727
## Positive -0.4208 [-0.9501; 0.1084] -1.56 0.1191
## Processing_Speed -0.1945 [-0.8642; 0.4752] -0.57 0.5692
## Total -0.5945 [-1.1076; -0.0814] -2.27 0.0232
## Verbal_Memory -0.0394 [-0.7091; 0.6303] -0.12 0.9081
##
## Quantifying heterogeneity / inconsistency:
## tau^2 = 0.0573; tau = 0.2394; I^2 = 35.8% [0.0%; 66.1%]
##
## Tests of heterogeneity (within designs) and inconsistency (between designs):
## Q d.f. p-value
## Total 20.26 13 0.0890
## Within designs 13.64 3 0.0034
## Between designs 6.62 10 0.7607
##
## Details of network meta-analysis methods:
## - Frequentist graph-theoretical approach
## - DerSimonian-Laird estimator for tau^2
## - Calculation of I^2 based on Q
Explanation:
For each modality, outcomes are compared within a network framework to
evaluate relative differences across clinical measures.
5. Individual Subgroup Meta‐Analyses
The following analyses focus on specific subgroups to illustrate the association between GSH levels and particular clinical outcomes.
5.1 Imaging – Total
res_imaging_total <- rma(yi = atanh(Correlation),
vi = 1/(SampleSize - 3),
data = imaging_total,
method = "REML")
cat("\n--- Imaging Total Meta-Analysis ---\n")##
## --- Imaging Total Meta-Analysis ---print(summary(res_imaging_total))##
## Random-Effects Model (k = 5; tau^2 estimator: REML)
##
## logLik deviance AIC BIC AICc
## -0.5464 1.0928 5.0928 3.8654 17.0928
##
## tau^2 (estimated amount of total heterogeneity): 0.0013 (SE = 0.0292)
## tau (square root of estimated tau^2 value): 0.0359
## I^2 (total heterogeneity / total variability): 2.64%
## H^2 (total variability / sampling variability): 1.03
##
## Test for Heterogeneity:
## Q(df = 4) = 5.0451, p-val = 0.2827
##
## Model Results:
##
## estimate se zval pval ci.lb ci.ub
## -0.1678 0.0911 -1.8419 0.0655 -0.3464 0.0108 .
##
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1# Forest plot
forest(res_imaging_total,
slab = imaging_total$Authors,
xlab = "Fisher's Z",
mlab = "Imaging Total")
# Funnel plot to assess publication bias
funnel(res_imaging_total, main = "Funnel Plot: Imaging Total")
Explanation:
This subgroup analysis examines overall GSH measures from brain imaging
studies and provides visual summaries (forest and funnel plots).
5.2 GSHt – Positive
res_gsht_positive <- rma(yi = atanh(Correlation),
vi = 1/(SampleSize - 3),
data = gsht_positive,
method = "REML")
cat("\n--- GSHt Positive Meta-Analysis ---\n")##
## --- GSHt Positive Meta-Analysis ---print(summary(res_gsht_positive))##
## Random-Effects Model (k = 14; tau^2 estimator: REML)
##
## logLik deviance AIC BIC AICc
## -0.3565 0.7130 4.7130 5.8429 5.9130
##
## tau^2 (estimated amount of total heterogeneity): 0.0261 (SE = 0.0210)
## tau (square root of estimated tau^2 value): 0.1616
## I^2 (total heterogeneity / total variability): 52.76%
## H^2 (total variability / sampling variability): 2.12
##
## Test for Heterogeneity:
## Q(df = 13) = 26.1734, p-val = 0.0161
##
## Model Results:
##
## estimate se zval pval ci.lb ci.ub
## -0.0070 0.0628 -0.1113 0.9114 -0.1300 0.1160
##
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1# Forest plot
forest(res_gsht_positive,
slab = gsht_positive$Authors,
xlab = "Fisher's Z",
mlab = "GSHt Positive")
# Funnel plot
funnel(res_gsht_positive, main = "Funnel Plot: GSHt Positive")
Explanation:
This analysis targets studies reporting blood total glutathione (GSHt)
in relation to positive symptoms.
5.3 GSHr – Negative
res_gshr_negative <- rma(yi = atanh(Correlation),
vi = 1/(SampleSize - 3),
data = gshr_negative,
method = "REML")
cat("\n--- GSHr Negative Meta-Analysis ---\n")##
## --- GSHr Negative Meta-Analysis ---print(summary(res_gshr_negative))##
## Random-Effects Model (k = 5; tau^2 estimator: REML)
##
## logLik deviance AIC BIC AICc
## 0.6333 -1.2666 2.7334 1.5060 14.7334
##
## tau^2 (estimated amount of total heterogeneity): 0.0286 (SE = 0.0315)
## tau (square root of estimated tau^2 value): 0.1692
## I^2 (total heterogeneity / total variability): 66.55%
## H^2 (total variability / sampling variability): 2.99
##
## Test for Heterogeneity:
## Q(df = 4) = 11.9360, p-val = 0.0178
##
## Model Results:
##
## estimate se zval pval ci.lb ci.ub
## -0.1007 0.0949 -1.0612 0.2886 -0.2867 0.0853
##
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1# Forest plot
forest(res_gshr_negative,
slab = gshr_negative$Authors,
xlab = "Fisher's Z",
mlab = "GSHr Negative")
# Funnel plot
funnel(res_gshr_negative, main = "Funnel Plot: GSHr Negative")
Explanation:
This subgroup analysis assesses the relationship between reduced
glutathione (GSHr) in blood and negative symptom severity.
Discussion
The results of this meta‐analysis indicate that while MRS‐based brain GSH measures do not show a robust association with clinical symptom severity, blood-based assessments—particularly of total glutathione (GSHt)—are significantly correlated with both symptom burden and cognitive performance. These observations lend support to the hypothesis that peripheral GSHt may serve as a biomarker for SSD severity. Methodological heterogeneity (e.g., differences in MRS protocols, cell types analyzed in blood assays) likely contributes to the observed discrepancies. Standardizing measurement techniques across studies is essential for clarifying the role of oxidative stress in SSD pathophysiology.
Extended Discussion
Recent research indicates that oxidative stress is linked not only to the presence of psychosis but also to its severity and functional outcomes. Markers of oxidative damage have been found to correlate with symptom severity in first-episode psychosis patients, including associations with more severe positive symptoms, greater negative symptom burden, and cognitive impairments. In particular, glutathione status (including reduced GSH, oxidized GSSG, and total GSH) may relate to these clinical features. For example, higher oxidized glutathione (GSSG) levels have been associated with worse cognitive performance in early-phase SZ, supporting the view that oxidative stress can contribute to the neurodegenerative aspects of the illness. Conversely, deficits in GSH-dependent defenses might manifest as or exacerbate negative symptoms; indeed, low GSH has been hypothesized to underlie certain negative symptom dimensions via NMDAR hypofunction and neuronal damage. Given that cognitive impairment and negative symptoms are major determinants of long-term disability in schizophrenia, uncovering links between GSH and these symptom domains is of high clinical interest.
Several studies have directly examined glutathione levels in patients and their relationship with symptom severity or cognitive function. However, findings have been mixed. Some report that lower GSH (in the brain or periphery) correlates with more severe overall symptoms or specific symptom domains, whereas others find no significant association. This inconsistency may stem from differences in illness stage, treatment status, or measurement techniques. To clarify the role of GSH in schizophrenia symptomatology, we performed a systematic review and meta-analysis of studies quantifying glutathione in patients with schizophrenia spectrum disorders and assessing its relationship to symptom severity or cognitive performance. In this report, we discuss our findings in the context of the broader literature, with explicit attention to symptom severity, cognitive function, and potential moderators such as brain region, illness phase, and antipsychotic treatment.
Several factors may explain discrepancies in the literature: - Illness Stage: First-episode or unmedicated patients tend to show stronger correlations between low GSH and increased symptom severity, whereas chronic patients (often medicated) may display more normalized GSH levels. - Regional Brain Differences: MRS studies indicate that regional variations (e.g., in the anterior cingulate cortex vs. medial temporal lobe) affect GSH measures, influencing their correlation with clinical outcomes. - Methodological Variations: Differences in measurement techniques, such as scanner field strength and spectral editing methods, contribute to variability in reported GSH values. - Patient Subtypes: Emerging evidence suggests the existence of distinct schizophrenia biotypes with respect to redox status, wherein only a subset of patients exhibits marked GSH deficits.
These insights underline the importance of standardized measurement and stratified analyses in future research.
Conclusion
This comprehensive meta‐analysis synthesizes evidence from both brain imaging and blood studies to examine the relationship between glutathione levels and clinical outcomes in schizophrenia spectrum disorders. Although brain GSH levels did not correlate significantly with symptom severity, the significant associations observed with blood GSHt levels underscore its potential as a biomarker. Future research should prioritize methodological standardization and explore the interplay between GSH, glutamatergic neurotransmission, and clinical outcomes.