Effectiveness of Chitosan as a Dietary Supplement in Lowering Cholesterol in Murine Models: A Meta-Analysis

This study presents a meta-analysis of studies that investigate the effectiveness of chitosan administration on lifestyle-related disease in murine models. A total of 34 published studies were used to evaluate the effect of chitosan supplementation. The effect sizes for various items after chitosan administration were evaluated using the standardized mean difference. Using Cochran’s Q test, the heterogeneity of effect sizes was assessed, after which a meta-ANOVA and -regression test was conducted to explain the heterogeneity of effect sizes using the mixed-effect model. Publication bias was performed using Egger’s linear regression test. Among the items evaluated, blood triglyceride and HDL-cholesterol showed the highest heterogeneity, respectively. Other than blood HDL-cholesterol, total cholesterol, and triglyceride in feces, most items evaluated showed a negative effect size with high significance in the fixed- and random-effect model (p < 0.0001). In the meta-ANOVA and -regression test, administering chitosan and resistant starch was revealed to be most effective in lowering body weight. In addition, chitosan supplementation proved to be an effective solution for serum TNF-α inhibition. In conclusion, chitosan has been shown to be somewhat useful in improving symptoms of lifestyle-related disease. Although there are some limitations in the results of this meta-analysis due to the limited number of animal experiments conducted, chitosan administration nevertheless shows promise in reducing the risk of cholesterol related metabolic disorder.


Introduction
Lifestyle-related diseases, including obesity, hyperlipidemia, atherosclerosis, type II diabetes, and hypertension, are widespread in industrialized countries, and are major threats to cardiovascular health. The syndrome is related to a combination of metabolic disorders, including abdominal obesity, hypertriglyceridemia, high-density lipoprotein (HDL) cholesterol decrease, hypertension, and high blood glucose, which lead to increased cardiovascular morbidity and mortality [1]. Unnatural blood lipid levels such as high levels of total cholesterol (TC) or triglyceride (TG), high low-density lipoprotein (LDL) level, or low HDL-cholesterol level are correlated with heart disease and stroke. Hypertension is one of the harmful risk factors for stroke and is a key factor in heart attacks. Moreover, obesity acts as a significant risk factor for cardiovascular disease and susceptibility to diabetes [2]. Thus, there has been an urgent need for effective methods of controlling these health-related parameters, including food additives.
Chitosan is one of the polymers containing acetyl glucosamine and glucosamine. It may be obtained by hydrolyzing and converting chitin with alkali from crabs, shrimps, insects, mushrooms, and the cell walls of microorganisms. Chitosan manufacture by deacetylation of chitin has been utilized in wastewater treatment and the agricultural sector. As the safety of chitin or chitosan has become increasingly recognized, it has recently-been used in a variety of fields, including medical supplies, food additives, and cosmetics [3,4]. Chitosan is also known among food additives of which the effects include lowering blood or liver cholesterol and triglyceride by combining with lipids [5]. It even shows an antiinflammatory effect by TNF-α inhibition [6][7][8][9]. Nauss et al. [10] assume that chitosan binds lipid micelle in the small intestine after the ingestion of a fatty meal, while Kanauchi et al. [11] propose a more specific mechanism by which chitosan inhibits fat digestion in the gastrointestinal tract. In the stomach, chitosan is dissolved in acidic gastric juice. In this aqueous phase, it acts as an emulsifier on fat globules. It also mixes with fat to form an emulsion. Once transferred into the intestine, the chitosan in the emulsion turns into an insoluble gel-like form trapped fat, which cannot be decomposed by enzymes such as pancreatin or other intestinal enzymes. As a result, fat excretion in feces is increased ( Figure 1). In this connection, [12] have confirmed that in one animal study chitosan administration led to fecal fat excretion approximately 7.5 times higher compared to that of a cellulose-fed group.
Mar. Drugs 2021, 19, x FOR PEER REVIEW 2 of 16 been used in a variety of fields, including medical supplies, food additives, and cosmetics [3,4]. Chitosan is also known among food additives of which the effects include lowering blood or liver cholesterol and triglyceride by combining with lipids [5]. It even shows an anti-inflammatory effect by TNF-α inhibition [6][7][8][9]. Nauss et al. [10] assume that chitosan binds lipid micelle in the small intestine after the ingestion of a fatty meal, while Kanauchi et al. [11] propose a more specific mechanism by which chitosan inhibits fat digestion in the gastrointestinal tract. In the stomach, chitosan is dissolved in acidic gastric juice. In this aqueous phase, it acts as an emulsifier on fat globules. It also mixes with fat to form an emulsion. Once transferred into the intestine, the chitosan in the emulsion turns into an insoluble gel-like form trapped fat, which cannot be decomposed by enzymes such as pancreatin or other intestinal enzymes. As a result, fat excretion in feces is increased (Figure 1). In this connection, [12] have confirmed that in one animal study chitosan administration led to fecal fat excretion approximately 7.5 times higher compared to that of a cellulose-fed group. Meta-analysis is a method of statistical analysis that combines results from various scientific studies to obtain a quantified synthesis [13]. Meta-analysis increases the power of statistical analysis by pooling the results from multiple available studies. Therefore, this study summarizes the results of various animal experiments and provides integrated technical data for clinical trials so that clinical trials can proceed more accurately.
Studies of lifestyle diseases in murine models suggest that they may be improved by administering chitosan. However, few comprehensive studies have been conducted to date on the effect of chitosan supplementation on improving lifestyle diseases. Accordingly, the objective of the present study was to perform a meta-analysis of the effects of chitosan on factors in lifestyle-related diseases in adults. Table 1 shows the data sets and experimental conditions for the 34 published studies used in the meta-analysis. The publication years of the studies ranged between 1978 and 2020. The animals most frequently used in the data set were rat strains such as Sprague-Dawley and Wistar, experiment duration was distributed between 2.8 and 21 weeks, and experimental diet most used for inducing hyperlipidemia in the data set was a high fat/cholesterol diet. In the case of Liu et al. [14], a high-fructose diet was used to induce hyperlipidemia. Furthermore, in the study of Gallaher et al. [15], blood total cholesterol Meta-analysis is a method of statistical analysis that combines results from various scientific studies to obtain a quantified synthesis [13]. Meta-analysis increases the power of statistical analysis by pooling the results from multiple available studies. Therefore, this study summarizes the results of various animal experiments and provides integrated technical data for clinical trials so that clinical trials can proceed more accurately.

Data Set
Studies of lifestyle diseases in murine models suggest that they may be improved by administering chitosan. However, few comprehensive studies have been conducted to date on the effect of chitosan supplementation on improving lifestyle diseases. Accordingly, the objective of the present study was to perform a meta-analysis of the effects of chitosan on factors in lifestyle-related diseases in adults. Table 1 shows the data sets and experimental conditions for the 34 published studies used in the meta-analysis. The publication years of the studies ranged between 1978 and 2020. The animals most frequently used in the data set were rat strains such as Sprague-Dawley and Wistar, experiment duration was distributed between 2.8 and 21 weeks, and experimental diet most used for inducing hyperlipidemia in the data set was a high fat/cholesterol diet. In the case of Liu et al. [14], a high-fructose diet was used to induce hyperlipidemia. Furthermore, in the study of Gallaher et al. [12], blood total cholesterol (TC) was observed in all studies. In addition to total triglyceride (TG), low-density lipoprotein (LDL)-and high-density lipoprotein (HDL)-cholesterol in the blood, TC and TG in the liver, and fecal TC and TG were investigated. The levels of chitosan administered to hyperlipidemia-induced animals ranged from 0.045 to 7.5% of the diet. The chitosan administration period varied between 3 and 21 weeks. Rat (Sprague-Dawley) 12 8

Effect Size and Heterogeneity
The effect sizes of chitosan administration on hyperlipidemia in murine models using fixed and random effect models are listed in Table 2. Most items other than HDLcholesterol in blood, total cholesterol, and triglyceride in feces showed negative effect size and high significance (p < 0.0001) in both effect models. These results mean that chitosan administration results in decreased levels of TC, TG, and LDL-C in blood, TC and TG in the liver, serum TNF-α and glucose in blood and body weight, and increased levels of blood HDL-C, fecal TC and TG.

Moderator Analysis
Since heterogeneity analysis in this study revealed a high level of heterogeneity between the studies analyzed, moderator analysis was performed to account for this. For this, meta-ANOVA and meta-regression were conducted. To perform the meta-ANOVA test, Q statistics between the subgroups (Q b ) calculated under assessing that between subgroups (τ 2 ) was the same. First of all, a meta-ANOVA analysis was performed on most items except fecal TG, as shown in Tables 3 and 4. Chitosan and resistant starch (CTS + RS) showed the highest effect size in blood TC and TG, body weight, blood glucose and blood HDL-C, CTS showed the largest effect size in blood LDL-C and TNF-α, and the cholestyramine (CSR) and water-soluble chitosan (WSC) showed the greatest effect size in liver TC and liver TG, respectively (Table 3). However, none of these items were statistically significant (p < 0.05). Table 4 shows the results of meta-ANOVA in analyzing the effect of chitosans administration period on biological indices (p > 0.05). Other than fecal TC, body weight, and blood glucose, most items showed significant differences (p < 0.05). In the case of TC, the Q statistics between the groups (Q b ) was 31.94 (df = 13, p = 0.0025); the effect size between groups was assumed to be significantly different.
Next, meta-regression was performed to evaluate the effect size between the type of chitosan used and the administration period (Table 5). Only CTS + RS was significant (p = 0.0208), and it was revealed to use to decrease blood TC. In the case of WSC, it was significantly effective in lowering of serum TNF-α and body weight (p = 0.0307 and 0.0008, respectively). With regard to the administration period, this was significantly relevant to blood HDL-C and liver TC with p = 0.0004 and 0.0358, respectively.

Publication Bias
Publication bias was conducted using an Egger's linear regression test (Table 6) on blood TC and TG, blood LDL-C and HDL-C, liver TG and TC, fecal TC, and body weight. As the results from the Egger's linear regression test show, significance was detected in all items (p < 0.05) indicating that the relationship between effect size and standard error was statistically significant and confirming the presence of bias [48]. Thus, the trim-and-fill technique was used to correct asymmetry due to publication bias in all items, with the resulting compensated effect sizes being shown in Table 7. Other than blood HDL-C, most of the effects showed significance (p < 0.05).

Discussion
In the results of Table 2, most items showed negative effect size and high significance (p < 0.0001) in both effect models. These results mean that chitosan administration results in decreased levels of TC, TG, and LDL-C in blood, TC and TG in the liver, TNF-α and glucose in blood and body weight, and increased levels of blood HDL-C, fecal TC and TG.
The bioavailability of dietary fat in the intestine decreased after chitosan administration. After this, reverse cholesterol transport, which is delivered from peripheral tissues to the liver, is accelerated by excretion of surplus dietary fat, resulting in an increase in the ratio of HDL-cholesterol [49]. Similarly, [50] have reported that the addition of chitosan to an animal diet caused a decrease in LDL-cholesterol content. Generally, HDL-cholesterol may decrease cardiovascular disease by converting cholesterol condensed on peripheral tissues or blood vessel walls into an ester compound. The ester compound is then transferred to the liver, excreted by bile-salt, and cholesterol content in blood is lowered. By contrast, LDL-cholesterol, which is the most general delivery type of blood cholesterol, accumulates easily on artery walls, causing arteriosclerosis. For this reason, it is known as the leading risk factor for arteriosclerosis and cardiovascular [51]. In this result, increased HDL-cholesterol, fecal total cholesterol, and triglyceride after chitosan administration are related to the factors mentioned above. According to Jeon and Kim [52], when chitosan is cationized (-NH 3 + ), its viscosity is increased by the formation of poly cations and gels. In high viscosity of the intestine, dietary fiber lower blood cholesterol by delaying cholesterol diffusion from micelle to mucosa, inhibiting bile acid metabolism, delaying micelle forming, and reducing cholesterol absorption rate in the intestine [19,53]. Based on this result, chitosan exhibits an excellent anti-hypercholesterolemic effect and is thought to be effective in mitigating cardiovascular disease caused by excessive fat intake.
Cytokines are secreted by activated lymphocytes and macrophages, and regulate the function of the cells related to immune response. They are also recognized as playing an essential role in the inflammatory response [54]. Yemak et al. [8] report that TNF-α generation was lower in lipopolysaccharide (LPS) and chitosan-injected mice than in LPSinjected mice. Similarly, Seo et al. [7] observed that TNF-α was increased by the application of special stimulants in a human mast cell line (HMC−1), but decreased by the use of chitosan. TNF-α is one of the pro-inflammatory cytokines synthesized by adipose tissue [55,56], and high TNF-α levels are one of the critical risk factors for diabetes [57]. In a similar vein, Yoon et al. [58] state that chitosan is associated with an anti-inflammatory response to TNF-α gene expression. According to Zhu et al. [59], chitosan has an anti-inflammatory effect on active molecules, for example TNF-α and IL-1β via the NF-κB pathway. Activated macrophages secrete numerous pro-inflammatory cytokines, including IL-1β and TNF-α, to intermediate the inflammatory response [60]. However, overproduction of these proinflammatory mediators causes excessive inflammation [61]; thus, regulation of the release of pro-inflammatory mediators may be important in mitigating the inflammatory response.
According to Prabu and Naturajan [62], blood glucose levels decreased in streptozotocininduced diabetic rats that were fed chitosan for 30 days. Other researchers suggest that the effectiveness of chitosan in lowering blood glucose may be due in part to the effect of total glyceride in lowering free fatty acids. Jo et al. [63] report that in an animal study, chitosan that was enzymatically treated and of low molecular weight (<1000 Da) was more effective in managing prandial glucose. Kim et al. [64] also report that chitosan that is low in molecular weight acted similarly to acarbose, a known anti-diabetic medication, in a murine model. They also note that chitosan administration inhibited sucrase and glucoamylase activities. It is recognized that chitosan binds with glucosidase in the intestinal brush border in a manner similar to acarbose (Hanefeld, [65]; Puls et al. [66]; Krentz and Bailey [67]). The inference of these reports is that body weight may be decreased by chitosan administration.
In the course of this process, heterogeneity is introduced as a result of methodological differences between studies. In general, a heterogeneity test is used to decide on methods for combining studies and to evaluate the consistency or inconsistency of findings (Petitti [68]; Higgins et al. [69]). To evaluate heterogeneity in relation to effect size in the present study, Q statistics and I 2 values were computed. The highest among Q statistics was TG in blood, with high significance (p < 0.0001). The significance of the Q statistic implies that the studies used to calculate the overall effect (the effect size of fixed and random effect models) do not share the same effect size with one another (Cho et al. [70]). In this study, the Q statistics for all items were found to be significant (p < 0.0001). However, one limitation of this method is its dependence on the number of studies (Fleiss [71]). I 2 and τ 2 values are commonly used to overcome this limitation of Q statistics by providing a concrete indication of heterogeneity. The I 2 value is used most frequently in meta-analysis to compare different numbers of studies and data types. Consequently, it offers a solution to the issue of the Q statistic when analyzing heterogeneity (Higgins et al. [72]). All items of I 2 value in the present study were above 70%, which means that they all showed significant levels of heterogeneity [73]. The τ 2 value indicates the absolute value of heterogeneity, representing variance in true effect sizes [74]. In addition, liver TG showed the highest τ 2 value, which means that variance in the effectiveness of chitosan administration is great (Cho et al. [70]).
Cholestyramine (trade name: Questran, Questran Light, Cholybar or Olestyr) and cholestipol (trade name: Colestid or Cholestabyl) as an anion-exchanger are these days used mainly for reducing cholesterol [75]. These medications contain amino groups, are water-insoluble, and unlike chitosan are not absorbed in the intestine. Specifically, they form insoluble complexes with bile acids in the intestines, which are then excreted in the feces. As a result, more plasma cholesterol is converted into bile acids in the liver to normalize its levels. When cholesterol is converted into bile acids, plasma cholesterol levels are lowered (National Institute of Diabetes and Digestive and Kidney Diseases [76]). Consequently, they are known to inhibit cholesterol absorption in the gut and to promote bile salt excretion. However, they are also known to involve a number of issues, including gastrointestinal disturbance, constipation, and colon cancer [77,78]. Valhouny et al. [79] report that chitosan supplementation showed a similar inhibition effect to cholestyramine in cholesterol adsorption. Similarly, an animal study by Jennings et al. [78] showed that chitosan was similar to cholestyramine in lowering lipids without other harmful changes in intestinal mucosa. Currently, a total of 1832 patents related to chitosan are being searched in the field of hyperlipidemia and associated cardiovascular diseases. It can thus be concluded that chitosan supplementation may be useful in lowering cholesterol and offers a promising alternative treatment for lifestyle-related diseases.

Data Set
To perform a meta-analysis of published studies regarding the effect of chitosan administration on lowering cholesterol in murine models between 1978 to 2020, a literature search was conducted on Pubmed (US National Library of Medicine, Bethesda, MD, USA) and Science Direct (Elsevier B. V., Amsterdam, The Netherlands). The keywords used for searching studies for meta-analysis were "chitosan, cholesterol" in all databases. The results obtained included 450 citations from Science Direct and 303 from Pubmed (US National Library of Medicine, Bethesda, MD, USA). These results were then filtered by title, abstract, and full text. Among them, 4 review articles and 7 studies of clinical tests in human studies were removed. Also, the studies expressed with graphical data were eliminated. Following this, studies regarding changes in cholesterol levels after chitosan administration were collected. Ultimately, a total of 34 studies with 11 items (e.g., total cholesterol, triglyceride, LDL-and HDL-cholesterol, TNF-α, and so on) were selected to perform a meta-analysis of the effectiveness of chitosan in reducing cholesterol in murine models.

Data Analysis
Corrected standardized mean difference (Hedges' g), and 95% confidence intervals (CI) were computed between control groups and treatment. The weight of the effect size was calculated using inverse-variance [80,81]. Effect-size analysis of fixed and random effect models was used to calculate overall effect due to differences in administration period, animal strain, and the type and dosage of chitosan used in each study. Cochran's Q test was performed to assess the statistical heterogeneity of the effect size, and the ratio of true heterogeneity to total variation in observed effects was expressed by the I 2 value. To confirm the heterogeneity of effect size using a mixed-effect model for the items in question, meta-ANOVA and regression analyses were also used. Meta-ANOVA and meta-regression analysis can evaluate the difference of Hedges' g among subgroups herein administration periods or type of treatment. The periods were set as independent factors in meta-ANOVA and as continuous variables in meta-regression. Finally, publication bias analysis was conducted to ensure the validity of the meta-analysis results. Statistical analysis and visualization of the results were performed using the 'meta', and 'metafor' packages in the R statistics software application (ver. 3.5.3, R Foundation for Statistical Computing, Vienna, Austria).

Conclusions
The present study confirmed the effectiveness of chitosan administration on lifestylerelated diseases through meta-analysis. Chitosan was significantly effective in lowering total cholesterol and triglyceride of blood and liver and rising fecal total cholesterol and triglyceride. Based on our results, chitosan was demonstrated to be useful in improving the symptoms of lifestyle-related disease.

Data Availability Statement:
The data presented in this study are fully available in the main text of this article.