Effects of Centella asiatica (L.) Urb. on cognitive function and mood related outcomes: A Systematic Review and Meta-analysis

Centella asiatica (L.) Urb. has been used as an herbal brain tonic for mental disorders and enhancing memory, but no review of the overall evidence of C. asiatica and cognitive function has been conducted. This study aims to determine the effects of C. asiatica on cognitive function and its related properties. The current systematic review includes five randomized controlled trials (RCTs) conducted to determine the effect of C. asiatica alone and six RCTs conducted to determine the effect of C. asiatica-containing products. Meta-analysis indicated that there are no significant differences in all cognitive function domains of C. asiatica when compared to placebo. However, it could improve mood by increasing alert scores [SMD: 0.71 (95% CI; 0.01 to 1.41); I2 = 30.5%] and decreasing anger scores at 1 hour after treatment [SMD: −0.81 (95%CI; −1.51 to −0.09); I2 = 36.6%]. None of the studies reported adverse effects of C. asiatica. In conclusion, there is not strong evidence to support the use of C. asiatica for cognitive function improvement in each cognitive domain. C. asiatica could improve alertness and relieve anger. However, some limitations should be aware including dose regimen, plant preparation, standardization, and product variation. Future well-designed clinical trials using suitable doses of standardized C. asiatica are still needed.


Results
Study selection. A total of 2,419 articles were identified from the database searches, and five articles were added based on our review of the reference lists. Of the articles, 693 were excluded because of duplication. A total of 1,785 titles and abstracts were screened. Of the screened titles and abstracts, 20 full-text articles were reviewed, of which only 11 articles were included in the systematic review. The flow of included studies is depicted in Fig. 1.

Study characteristics.
Of the 11 included studies, five studies (45%) compared C. asiatica alone to placebo, and six studies (54%) compared a combination of C. asiatica versus other herbs. For combination products, three of the six studies (50%) used mix herbs as the active ingredients, two of the six studies (33%) used Gingko biloba as the major compound, and one study used a combination of vitamins and herbs (Table 1). Only one study (9%) did not report the Latin binomial nomenclature of the herbal ingredients 27 . Standardization methods were reported in three studies (27%) 26,28,29 but only two studies quantitatively described the standardization 26,29 . Nine studies (81%) were conducted using double-blind parallel designs, one used an open-labeled parallel design, and one used a cross-over design. Most studies (91%) were conducted in healthy volunteers, while one study was conducted in children with attention deficit hyperactive disorder. Other information (herbal supplement type, dosage form, plant preparation, dose of C. asiatica, standardization method, study characteristics, intervention and patient characteristics) is summarized in Table 1.
Quality of included studies. Three of the studies (27%) [29][30][31] had a high risk of bias, seven studies (64%) were unclear 25,27,28,[32][33][34][35] , and one study (9%) had a low risk of bias 26 . Although, all studies stated that they were randomized controlled trials, four of the trials (36%) were found to have unclear risk of bias for "sequence generation" because there was no description of the sequence generation methods. Most studies (72.7%) did not describe the "allocation concealment" method. In the bias domain of "blinding", one study was an open-label study which was categorized as having a high risk of bias. All double-blind studies included had low risk. Furthermore, incomplete outcome data, selective outcome reporting, other sources of bias risk, and JADAD scores for each study are presented in Table 2.

Effects of C. asiatica in cognitive function.
Of the included studies, 60 cognitive function tests were described, but only 27 of the tests had sufficient data for a meta-analysis. The 27 cognitive function tests were each categorized into specific cognitive domains for the purpose of evaluating the cognitive improvement effect of C. asiatica 1 . The domains included 1) overall cognitive status, 2) attention and concentration, 3) executive function, 4) working memory, 5) information processing speed, 6) language, 7) verbal memory, 8) visuospatial skill, and 9) visual memory ( Table 3). The meta-analysis indicated no significant difference between C. asiatica  Table 4. However, the findings in some trials indicated that C. asiatica alone may improve working memory. Significant positive effects were found on numeric working memory tests (Appendix D) (i.e., decreased working time) after patients received 750 mg ( Adverse effects. Adverse effects or toxicity associated with C. asiatica were also evaluated based on the included articles. No adverse effects were reported in any studies looking at C. asiatica alone. However, for studies of combination products, four studies reported mild adverse events of C. asiatica-containing products. Two studies reported adverse event rates comparable to the placebo rate 31,34 , while another two studies reported lower rates of adverse event for C. asiatica-containing products 30,33 . Common adverse events were gastrointestinal discomfort, flatulence, nausea, headache, decreased appetite, sedation, and rash. Hepatotoxicity, which has been reported in one previous case report 36 , was not observed in any of the included RCTs.

Discussion
This systematic review and meta-analysis provides a comprehensive summary of the effects of C. asiatica on cognitive function.
Current evidence does not support the effects of C. asiatica alone on overall cognitive function. However, ingestion of C. asiatica water extract (750 mg/day) for 1 hour may improve working memory, as shown in the positive effect on the numeric working memory test 26 by a decrease in working time. This finding does not agree with a recent quasi-experimental study which found a statistically significant improvement in the memory domain of patients who had vascular cognitive impairment treated with C. asiatica extract (1,000 mg/day) when compared  to patients treated with 3 mg/day of folic acid 37 . In that study, however, the dose of C. asiatica was higher than nine of the eleven trials included in this meta-analysis. Thus, the non-significant differences in overall cognitive function between C. asiatica and its comparators observed in this review might be due to the dosages used in the included studies. In traditional use and experimental evidence 38 , at least 3 grams of C. asiatica needs to be used to improve cognitive function. However, only two included studies 29, 32 used doses greater than 3 g of C. asiatica per day, while the rest used lower doses.
The combination of C. asiatica with other herbs also showed non-significant improvements in overall cognitive function. However, the combination products in other studies have revealed that there arepossible effects on attention and concentration 30 , executive function 34 and information processing speed 30 . The improvement in cognitive function from the combination products might be due to the synergistic effects of C. asiatica with other herbs or the effects of other herbs in C. asiatica-containing products such as G. biloba 31,34 . G. biloba is a well-known herbal medicine used for cognitive impairment. From previous systematic reviews and meta-analyses, G. biloba exhibited potential benefits for cognitive improvement in mild cognitive impairment or Alzheimer's patients 2,5 . Moreover, beneficial effects on cognitive function of Withania somnifera (L.) Dunal, Spirulina (Arthrospira platensis) and paeoniflorin (monoterpene glucoside) have been exhibited in different pre-clinical models 30 . None of the studies reported details on which parts of C. asiatica were used in the combination or how the combinations were prepared. Thus, thefindings could not show the direct effect of C. asiatica on cognitive function, and there is currently a lack of persuasive evidence to confirm a cognitive enhancing effect of C. asiatica.
For secondary outcomes, C. asiatica consumption was associated with improvements in self-reported alertness (after 2 months of ingestion) and with reductions in self-reported anger (after 1 hour of ingestion). Moreover, C. asiatica alone (750 mg/day for 2 months) induced alertness and calmness. These improvements in alertness and calmness may facilitate cognitive function by improving working memory, attention and concentration, executive function and information processing speed, and memory capacity and by reducing the time to solve problems. These results also support the traditional use of C. asiatica as a brain tonic. However, the positive effects may be caused by the other herbs in the combination products, so firm conclusions on the efficacy of C. asiatica cannot be drawn. There were also no significant differences between C. asiatica and placebo for physical or total QoL   scores. From the safety data, C. asiatica seems to be safe since there were no serious adverse events reported in any of the included articles. This meta-analysis included both C. asiatica alone and C. asiatica combined with other herbs. There were differences among the included studies such as differences in the part of C. asiatica used, dosage forms, extraction procedures, preparation, and outcome measurements. However, based on the objectives, all cognitive function data were collected from the RCTs that used any type of C. asiatica. The authors believe that the analysis is valid to address the objectives. Using the standardized mean difference (SMD), allowed the effect of C. asiatica on cognitive function to be assessed across the various types of cognitive function measurements used in the included studies. SMD converts data from different scales to a common scale. However, the standardization causes the original information for each measurement to be lost, so the findings cannot be interpreted in common units. They can only provide the level of significance of the effect of C. asiatica compared to the comparators 39 .
A classification defined by previous studies was used to determine the domains of cognitive function and pool the findings together 1,37 . This classification has been used in several studies 37,[40][41][42] to classify the domains of cognitive function and pool their findings. Thus, it is believed that the approach is appropriate for this meta-analysis. As no validation study of the classification was conducted, future studies may look into this issue.  This review identified limitations in the reporting of clinical studies of C. asiatica. Most of the included studies did not report details on the parts of the C. asiatica plant used in the products, the standardization methods, the active marker contents, or the methods for preparing the products. Only three of eleven (27%) trials 26,28,30 reported standardization methods of the plant extract, and only two trials (18%) reported the amount of the active compounds (asiaticoside and asiatic acid) contained in the C. asiatica extract 26,28 . Moreover, none of the studies reported qualitative analyses (such as HPLC fingerprints) of the C. asiatica in their studies. Clinical trials of herbal medicine should use standardized products as interventions and should report the detail of each intervention according to the CONSORT statement for reporting herbal medicinal interventions 43 . Furthermore, the place, conditions, and season of cultivation as well as the parts of the plant used can affect the pentacyclic triterpene (asiaticoside, asiatic acid, madecassoside, madecassic acid) contents of the C. asiatica raw material 44 . Lack of herbal standardization in clinical trials may affect the quality of studies and explain the variations in the clinical effects across studies. Interpretation of the findings of this systematic review should be done with cautions due to the lack of information about standardization.
Another consideration is that the doses of C. asiatica in each study were different, ranging from 40-12,000 mg/ day. Variations in C. asiatica preparation were also observed. For C. asiatica alone, three of five trials used dry C. asiatica powder ranging from 500-12,000 mg/day while two trials used C. asiatica water extract ranging from 250-750 mg/day. Furthermore, the doses used in most of the included studies were lower than the traditional dose recommendation for cognitive improvement (3 g/day of C. asiatica powder) 38 . In the combination products, the dose of C. asiatica was very low (40-204 mg/day) compared with the main active component except in one study that used C. asiatica 1,000 mg/day 29 . Additionally, the dose and preparation of C. asiatica in some combination products was not clear. These limitations may affect the pooled data of C. asiatica in each cognitive domain. Moreover, the observed findings did not support a direct effect of C. asiatica containing products on cognitive function. There is currently a lack of persuasive evidence to confirm a cognitive enhancing effect of C. asiatica.
Based on this review, future well-designed clinical trials are warranted to evaluate the effects of C. asiatica products on cognitive function and mood as well as its safety. Standardized doses of C. asiatica products should  Table 4. Result of primary and secondary outcomes. * Significant (p < 0.05). #Presented as mean difference (not standardized mean difference). Combination only = Only combination product, C. asiatica only = C. asiatica alone product. Outcomes at the end of study = Outcomes measured at the longest following up. *Significant (p < 0.05), CI = confident interval. #Presented as mean difference (not standardized mean difference). All = pooled all data, Combination only = Only combination product, C. asiatica only = C. asiatica alone product, DB = Double blind, Score = Score unit, Time = Time unit, Healthy = Healthy volunteer, Elderly = elderly volunteer. Outcomes at the end of study = Outcomes measured at the longest following up time.
Scientific RepoRts | 7: 10646 | DOI:10.1038/s41598-017-09823-9 be investigated over short-term and long-term periods of ingestion for effects in each specific cognitive domain, especially working memory, attention and concentration, executive function, and information processing speed.
In conclusion, the findings revealed that there is no strong evidence to support the effect of C. asiatica on overall cognitive function improvement. However, C. asiatica may improve working memory. A combination of C. asiatica with other herbs may improve attention and concentration, executive function, and information processing speed. C. asiatica may also improve mood disorders in terms of self-reported alertness and reductions in self-reported anger. Issues with dosage and preparation standardization need to be considered when these findings are applied. Future well-designed clinical trials are needed to assess the effects of standardized C. asiatica on cognitive function and mood as well as safety.

Methods
This systematic review was conducted according to the Cochrane Collaboration framework guidelines 39 and was reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) Statement 45 . The review protocol was registered with PROSPERO (registration number: CRD42015023595).
Search strategies and study selection. An electronic search was conducted for original articles from inception to September 2016 using a number of electronic databases including AMED, CINAHL, Cochrane Central Register of clinical trial, EMBASE, PubMed, Psycinfo, Science direct, Scopus, www.clinicaltrials.gov, ThaiLis, Thai Index Medicus, Thai Medical Index, and Thai Thesis Database. Strategic search terms were C. asiatica name, OR active compound from C. asiatica (such as asiaticoside, madecassoside, asiatic acid, madecassic acid), OR C. asiatica containing products combined with cognitive function or memory and its related properties including mood and quality of life. Details of the search strategies are described in appendix A. Eligibility criteria were 1) published and unpublished randomized controlled trials in patients or healthy volunteers and 2) reported effects of C. asiatica or a combination of C. asiatica with other herbs in humans. No language restriction was applied. To ensure that the search would be thorough, reference lists were reviewed to identify potential studies not indexed in above mentioned databases. Furthermore, corresponding authors of identified studies were consulted for additional studies as sources. Titles and abstracts were screened according to the eligibility criteria. Full-text articles of the potential studies were retrieved from database or corresponding authors and were subsequently assessed independently by two researchers (PP, PD) for inclusion in the meta-analysis. Disagreements between the independent researchers were settled by discussion and consensus with a third independent researcher (NC).
Data extraction and quality assessment. Data extraction was undertaken using a standard data extraction form. Extracted data included study design, characteristics of participants, characteristics of intervention and comparator, duration of herbal use, follow-up time, cognitive function tests, and cognitive function outcomes. Data for cognitive function tests included the name of the cognitive function test, the cognitive function domain, the outcome measures, and the outcome scale. For this meta-analysis, each cognitive test was categorized into one specific domain of cognitive function following a previous study 1 . This approach avoids over-weighting effects and provides consistency for the evaluation of the effect of C. asiatica on cognition across studies (Appendix B, C). A primary outcome of interest was the clinical effect of C. asiatica on cognitive function in each domain (Appendix B) including attention and concentration, executive function, information processing speed, language, visuospatial skill, working memory, verbal memory, and visual memory as well as overall cognitive status. In addition, secondary outcomes were mood, quality of life, and adverse events reported across each intervention. Where relevant data were unavailable, it was sought directly from the corresponding authors.
The quality of included studies was assessed using the Cochrane risk of bias tool 39 and JADAD score 46 . Sequence generation, allocation concealment, blinding, incomplete outcome data, selective outcome reporting, and other sources of bias were evaluated. Data search, data extraction, and quality assessment were performed by PP and PD. Disagreements between the reviewers were settled through discussion and consensus.

Statistical analysis.
To determine the cognitive effect of C. asiatica, data for individual cognitive function tests were compared between C. asiatica and its comparator using standardized mean difference (SMD) or mean difference (MD) with a 95% confidence interval (CI). Heterogeneity was assessed by the I 2 -statistic 47 . Thresholds of I 2 were interpreted in accordance with the magnitude and direction of effects and strength of evidence of heterogeneity. I 2 values of more than 50% indicated substantial heterogeneity. Data from included studies were pooled using the Der Simonian and Laird random-effects model 48 . The software used for data analysis was STATA version 12 (STATA Corp, College Station, TX, USA). Data availability. The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.