Association of reduced glutathione levels with Plasmodium falciparum and Plasmodium vivax malaria: a systematic review and meta-analysis

Reduced glutathione (GSH) is a crucial antioxidant with recognized roles in malaria pathogenesis and host response. Despite its importance, reports on the association of GSH with malaria are inconsistent. Therefore, this systematic review and meta-analysis investigated the differences in GSH levels in relation to Plasmodium infection. A comprehensive literature search of six electronic databases (Embase, MEDLINE, Ovid, PubMed, Scopus, and ProQuest) was conducted. Of the 2158 initially identified records, 18 met the eligibility criteria. The majority of studies reported a significant decrease in GSH levels in malaria patients compared with uninfected controls, and this was confirmed by meta-analysis (P < 0.01, Hedges g: − 1.47, 95% confidence interval [CI] − 2.48 to − 0.46, I2: 99.12%, 17 studies). Additionally, there was no significant difference in GSH levels between Plasmodium falciparum malaria and P. vivax malaria (P = 0.80, Hedges g:  0.11, 95% CI − 0.76 to 0.98, I2: 93.23%, three studies). Similarly, no significant variation was observed between symptomatic and asymptomatic malaria cases (P = 0.78, Hedges g: 0.06, 95% CI − 0.34 to 0.46, I2: 48.07%, two studies). In conclusion, although GSH levels appear to be generally lower in malaria patients, further detailed studies are necessary to fully elucidate this complex relationship.


Search strategy
A comprehensive literature search was conducted across six electronic databases (Embase, MEDLINE, Ovid, PubMed, Scopus, and ProQuest).The search terms included "("reduced glutathione" OR GSH OR "gamma-L-Glutamyl-L-Cysteinylglycine" OR "gamma L Glutamyl L Cysteinylglycine" OR "gamma-L-Glu-L-Cys-Gly" OR "gamma L Glu L Cys Gly") AND (malaria OR plasmodium OR "Remittent Fever" OR "Marsh Fever" OR "Paludism)", with slight variations for each database (Supplementary Table S1).We also searched Google Scholar and the reference lists of selected articles.Although Google Scholar is an expansive database, it was used as a supplementary literature search tool.This decision was based on its limitations: it only supports basic Boolean operators in search strings, lacks the ability to export results in bulk as citations, and displays only the first 1000 search records, which cannot be sorted 23 .Additionally, only the first 200 search records were screened for eligibility, as previously suggested 23 .The searches were restricted to articles written in English, but there was no restriction on the publication date.The searches began from database inception to June 12, 2023.

Study selection and eligibility criteria
The study selection process was conducted in a stepwise manner by two independent authors (M.K. and A.M.).Any disagreement between the two authors was resolved via consensus.
First, duplicate records from the various database searches were removed manually and using automated tools.Then, the remaining unique records underwent a screening process, where irrelevant studies and those without abstracts were excluded.Next, the full texts of the remaining articles were retrieved for a detailed eligibility assessment.Studies were included in the review if they were original research articles, reported GSH levels in malaria and uninfected controls, and provided sufficient quantitative data for meta-analysis, such as the mean (with standard deviation) or median (with interquartile range) GSH levels 24 .The exclusion criteria encompassed in vitro/in vivo studies, review articles, studies lacking GSH information, studies not specifying the GSH type, studies without malaria cases, conference abstracts, or studies analyzing post-treatment GSH levels.

Data extraction and quality assessment
The following data were extracted for each eligible study: first author's name, year of publication, study design, year in which the study was conducted, geographic location, targeted Plasmodium species, clinical status, data on GSH levels, method of malaria detection, and method of GSH measurement.The quality of the included studies was assessed using Joanna Briggs Institute Critical Appraisal tools, depending on the study design (crosssectional, cohort, or case-control) 25 .The cross-sectional studies were evaluated for clarity of inclusion criteria, validity of exposure and outcome measurements, and management of confounding factors.The cohort studies were appraised based on group similarity, validity of exposure measurement, strategies for handling confounding factors, adequacy of the follow-up period, and appropriateness of statistical analyses.The case-control studies were assessed for group comparability, appropriateness of case-control matching, validity of exposure measurement, management of confounding factors, and sufficiency of the exposure period.Responses were categorized as "Yes, " "No, " "Unclear, " or "Not applicable" based on the relevance and availability of information for each criterion.The quality rank of an individual study was determined by the percentage of "Yes" responses among all items as follows: > 75th percentile, high quality; 50th-75th percentile, moderate quality; and < 50th percentile, low quality 26 .

Data synthesis and statistical analysis
The extracted data were used for qualitative synthesis.Additionally, a meta-analysis was conducted for quantitative synthesis.For the meta-analysis, the standardized mean difference (Hedges g) of GSH levels between groups of participants was calculated along with their 95% confidence intervals (CIs).Heterogeneity was quantified using the I 2 statistic 27 as follows: 0-40%, low heterogeneity; 30-60%, moderate heterogeneity; 50-90%, substantial heterogeneity; and 75-100%, considerable heterogeneity 27 .To explore the potential sources of heterogeneity, a meta-regression analysis using various factors, like the publication year, study design, continent, participant groups, Plasmodium species, diagnostic method for malaria, and quality rank of included studies, was conducted 28 .Subgroup analysis was conducted based on publication year, study design, geographic location, participant group, Plasmodium species, diagnostic method for malaria, and quality rank of included studies.A sensitivity analysis was performed using the fixed-effect model and the leave-one-out meta-analysis 29 .The leaveone-out meta-analysis was used to determine the effect of each individual study on the pooled effect estimate of the remainder of the studies 29 .Publication bias was evaluated by funnel plot analysis and Egger regression test 30 .All statistical analyses were performed using Stata v17.0 software (StataCorp.College Station, TX).P-values < 0.05 were considered statistically significant.

Characteristics of the included studies
The 18 included studies showed diverse characteristics.The majority were published during 2010-2023 (83.3%), indicating a greater interest in the investigation of the association between oxidative stress and malaria during this period.Of the studies included, only 50% specified the year in which they were conducted.The research  The common method of malaria detection was light microscopy (72.2%) (Tables 1, 2, Supplementary Table S2).

Quality of the included studies
For the cross-sectional studies, the majority 15,16,20,32,35,40 exhibited robust adherence to the predetermined criteria but failed to address confounding factors.However, three studies 17,19,38 successfully met all quality parameters.Two studies 36,41 , although fulfilling most criteria and identifying confounding factors, neglected to proffer strategies to mitigate confounders.Regarding the cohort studies, only one 34 met all quality parameters.However, the other cohort study 39 displayed certain inadequacies and ambiguities, including exposure measurement and handling of confounding factors.For the case-control studies, one 14 struggled with group comparability and case-control matching and did not address confounding factors.Another 31 upheld all quality criteria.The remaining three case-control studies 18,33,37 , while fulfilling most criteria, were vague regarding case-control matching and the duration of the exposure period (Supplementary Table S3).

Qualitative synthesis
The majority of African studies on GSH levels in malaria were conducted in Nigeria, with significant contributions from authors such as Abduljalil et al. 14 , Abubakar et al. 15 , Akanbi et al. 16 , Babalola et al. 19 , Nsonwu-Anyanwu et al. 37 , Oluba et al. 39 , Onyeneke et al. 40 , and Ozojiofor et al. 41 .The majority of Asian studies were conducted in India, as evidenced by works from Aqeel et al. 31 , Bhattacharya and Swarup-Mitra 33 , Das and Nanda 34 , Sohail www.nature.com/scientificreports/et al. 17 , and Tyagi et al. 18 .Overall, a significant decrease in GSH levels in individuals with malaria compared with uninfected controls was observed in 13 of these studies.Regarding different age groups, the majority of studies focused on children observed a significant decrease in GSH levels when infected with malaria, as seen in studies by Abduljalil et al. 14 , Abubakar et al. 15 , Das and Nanda 34 , Oluba et al. 39 , and Ojongnkpot et al. 38 .In particular, Ojongnkpot's study from Cameroon highlighted a negative relationship between GSH levels and parasite density in children 38 37 .In Indonesia, Fitri et al. 20 noted no difference in GSH levels between severe and nonsevere malaria in adults.Two Nigerian studies specifically focused on pregnant and nonpregnant women: Akanbi et al. 16 and Onyeneke et al. 40 .Both reported decreased GSH levels in subjects with malaria compared with uninfected controls.However, Onyeneke et al. did not find an association between GSH levels and parasite density in this group 40 .Some studies enrolled participants from all age groups, with Bhattacharya and Swarup-Mitra 33 and Ozojiofor et al. 41 , from India and Nigeria, respectively, noting a decline in GSH levels in malaria patients.In contrast, the Indian studies by Sohail et al. 17 and Tyagi et al. 18 observed increased GSH levels in malaria patients, providing an intriguing counterpoint.

GSH levels between P. falciparum and P. vivax
Three studies investigated GSH levels in both P. falciparum and P. vivax malaria cases, providing quantitative data applicable to the meta-analysis 17,31,36 .The results showed a significant difference in GSH levels between P. falciparum and P. vivax malaria (P = 0.80, Hedges g: 0.11, 95% CI: − 0.76 to 0.98, I 2 : 93.23%, three studies; Fig. 3).Performing both meta-regression and subgroup analysis was not feasible in this context due to the restricted number of studies available.

GSH levels between symptomatic and asymptomatic malaria
Two studies investigated GSH levels in both symptomatic and asymptomatic P. falciparum malaria and reported quantitative data, which we used in the meta-analysis 19,38 .The results revealed no significant difference in GSH levels between symptomatic and asymptomatic malaria cases (P = 0.78, Hedges g: 0.06, 95% CI: − 0.34 to 0.46, I 2 : 48.07%, two studies; Fig. 4).Due to the limited number of studies available, both a meta-regression and subgroup analysis could not be conducted.

Sensitivity analysis
Two approaches were employed to perform sensitivity analysis: the fixed-effect model and the leave-one-out meta-analysis.The fixed-effect model revealed a significant reduction in GSH levels in individuals with malaria compared with uninfected controls (P = 0.01, Hedges g: − 0.12, 95% CI − 0.21 to − 0.03), as confirmed by the meta-analysis of 17 studies (I 2 : 99.12%, Supplementary Fig. S1).The leave-one-out meta-analysis pinpointed the study by Bhattacharya and Swarup-Mitra 33 as an outlier.Its removal altered the meta-analysis results (P = 0.052, Hedges g: − 0.99, 95% CI − 1.98 to − 0.01; Fig. 5).

Publication bias
Two standard methodologies were implemented to evaluate the presence of publication bias: a funnel plot analysis and Egger regression test.The results depicted in the funnel plot were asymmetrical (Fig. 6), suggesting an imbalanced distribution of studies around the mean effect size, implying substantial publication bias.Egger test was further conducted to quantify the bias captured in the funnel plot.Notably, this test identified a significant result (P < 0.01).Therefore, both the asymmetrical funnel plot and the significant Egger test result collectively indicate the possible presence of publication bias due to the small-study effect in the meta-analysis.

Discussion
Qualitatively, most studies concur that GSH levels significantly decrease in malaria patients compared with uninfected controls, corroborating the role of GSH in malaria pathogenesis.Furthermore, the meta-analysis confirmed that GSH levels significantly decreased in malaria patients compared with uninfected controls.GSH, a critical antioxidant in human cells, protects the body from damage caused by oxidative stress 42 .The observed decrease in GSH levels in malaria patients may be a consequence of the body utilizing its GSH reserves to counteract the oxidative stress caused by malaria.Reductions in host GSH levels during malaria can have dual implications for disease pathogenesis.On the one hand, diminished GSH renders host erythrocytes more vulnerable to oxidative damage, potentially exacerbating disease symptoms due to increased oxidative stress 43 .Conversely,   www.nature.com/scientificreports/ the malaria parasite, particularly P. falciparum, relies on host GSH to detoxify and resist antimalarial drugs 4 .Therefore, while reduced GSH may weaken the host's defense against oxidative damage, it can also hinder the parasite's ability to survive drug treatments.Thus, the intricate balance between host and parasite GSH dynamics underscores the complex nature of malaria pathogenesis 44 .However, this generalized trend is subject to certain exceptions.For instance, Aqeel et al. found decreased GSH levels, specifically in patients with P. vivax malaria but not in patients with P. falciparum malaria 31 .This divergence may stem from the different pathophysiological mechanisms employed by P. falciparum and P. vivax.Since P. vivax and P. falciparum exhibit different levels of disease severity, this may influence the level of oxidative stress and, consequently, GSH consumption 45 .For example, infection with P. falciparum can lead to more severe complications, yet some P. vivax cases can also develop severe malaria 46,47 .Nevertheless, no difference in GSH levels between P. falciparum and P. vivax malaria was observed, suggesting that more studies are necessary to determine differences in the distinct antioxidant levels between these two species.Furthermore, the meta-analysis of the clinical status of patients revealed no significant difference in GSH levels between symptomatic and asymptomatic malaria cases.This suggests that alterations in GSH levels are consistent, irrespective of the presence of symptoms.Additionally, a persistent reduction of plasma GSH levels in the early stages of the disease has been observed, which was later compensated during the advanced phase 31 .Thus, reduced GSH levels may be common during the acute phase of Plasmodium infection, regardless of disease severity or whether the infection is caused by P. vivax or P. falciparum.
Other subgroup meta-analyses demonstrated differences in GSH levels between malaria patients and controls across specific parameters, such as geographic location and participant groups.Notably, African studies showed a significant difference in GSH levels between malaria patients and controls, whereas Asian studies did not.This discrepancy may be because the African studies mainly enrolled children.In the subgroup analysis of age groups of participants, only studies involving children showed significant differences in GSH levels, possibly indicating that malaria has a more pronounced effect on GSH levels in this age group.Furthermore, the less mature immune systems of children may impact the oxidative stress response 48 .Additionally, children in Africa have lower antioxidant levels, which may lead to more severe malaria cases 49 .
Interestingly, Babalola et al. 19 and Fitri et al. 20 reported no significant differences in GSH levels between different patient groups.This finding may be attributed to a range of factors, including the timing of sample collection, the disease stage, and individual variations in patients' antioxidant response.In contrast, two studies reported increased GSH levels in malaria patients 17,18 .While this may initially seem counterintuitive given the body's typical response to oxidative stress, it is important to note that the elevated GSH levels are likely attributed to the parasite itself.In their bid to survive and resist antimalarial drugs, Plasmodium species can upregulate GSH synthesis, leading to observed increases in overall GSH levels within the host 50,51 .Sohail et al. proposed that the observed increase in GSH levels in malaria patients could be due to transitional polymorphisms within GSTs, which might enhance the host's GSH availability 17 .Tyagi et al. suggested that increased GSH levels among malaria patients might be due to decreased GSH utilization 18 .
The relationship between GSH levels and parasite density varied across studies, with some reporting a negative relationship and others finding no significant association 33,38 .These discrepancies may be attributed to factors such as variations in the host's immune response, the parasite's lifecycle stage when the sample was collected, or differences in Plasmodium species.Oxidative stress and sickle cell disease are related to each other 52 .One of the included studies demonstrated that patients with malaria and sickle cell disease experienced severe oxidative stress 32 .The authors reported that although the GSH levels were higher in patients without sickle cell disease compared with those with malaria and sickle cell disease, the difference was not statistically significant.Therefore, patients with both malaria and sickle cell disease may have a higher demand for GSH to detoxify the increased oxidative stress 32 .
While the meta-analysis demonstrated a significant reduction in GSH levels in malaria patients compared with uninfected controls, indicating an association between malaria and lower levels of this critical antioxidant, the high variability among studies (I 2 : 99.12%) must not be ignored, suggesting substantial heterogeneity in the results.A limited number of studies investigated GSH levels in various contexts: (i) malaria patients with severe complications versus those without severe complications, (ii) asymptomatic versus symptomatic malaria, and (iii) P. falciparum versus non-P.falciparum malaria.Thus, the conclusions of our study were limited.Among the high-quality studies, there was a significant difference in GSH levels with a substantial effect size.Conversely, moderate-quality studies did not show a statistically significant difference.This may highlight the importance of study quality when interpreting results.Additionally, the level of heterogeneity remained high when the subgroup analyses were performed.Therefore, the true confounders of the relationship between GSH and malaria remain unidentified.Other potential confounders may include both infectious and noninfectious conditions that are co-endemic with malaria, such as nutritional deficiencies 53 , diabetes 54 , human immunodeficiency virus 55 , and coronavirus disease 2019 56 .Based on the information from the included studies, the timing of sample collection was not explicitly stated in each study.Consequently, a meta-regression analysis to test whether the timing of sample collection influenced the effect estimate of the meta-analysis could not be performed.The final limitation is the evidence of significant publication bias, and the sensitivity analysis calls for a cautious interpretation of these findings.The considerable impact of a single study on the overall meta-analysis outcome underscores the importance of incorporating a diverse range of studies to mitigate potential biases.

Conclusion
This comprehensive review and meta-analysis of the existing literature indicate a trend of decreased GSH levels in malaria patients compared with uninfected controls, which is consistent with the majority of the reviewed studies.Furthermore, the meta-analysis underlines the potential of GSH as a diagnostic biomarker for malaria.However, the relationship between GSH levels and specific characteristics, such as Plasmodium species, malaria

Figure 1 .
Figure 1.Flow diagram of the study selection process.

Figure 2 .
Figure 2. Forest plot showing the difference in the reduced glutathione levels between malaria patients and uninfected controls.CI confidence interval, N number of participants, SD standard deviation, blue square effect estimate, green diamond pooled effect estimate, red vertical line pooled effect estimate, gray vertical line no effect line.

Figure 3 .
Figure 3. Forest plot showing the difference in the reduced glutathione levels between patients with P. falciparum malaria and those with P. vivax malaria.CI confidence interval, N number of participants, SD standard deviation, blue square effect estimate, green diamond pooled effect estimate, red vertical line pooled effect estimate, gray vertical line no effect line.

Figure 4 .Figure 5 .
Figure 4. Forest plot showing the difference in the reduced glutathione levels between patients with symptomatic and asymptomatic malaria.CI confidence interval, N number of participants, SD standard deviation, blue square effect estimate, green diamond pooled effect estimate, red vertical line pooled effect estimate, gray vertical line no effect line.

Figure 6 .
Figure 6.Funnel plot showing an asymmetrical distribution of the effect estimates between the middle line of the plot.CI confidence interval, Blue dot effect estimate, red vertical line pooled effect estimate.

Table 2 .
Characteristics of each study.GSH reduced glutathione, RDT rapid diagnostic test.

Table 3 .
Subgroup analyses of GSH levels between malaria cases and uninfected controls.CI confidence interval, N/A not assessed, RDT rapid diagnostic test.