Effect of onion on blood lipid profile: A meta‐analysis of randomized controlled trials

Abstract Background Studies indicate that onion supplementation may be effective in the treatment of dyslipidemia; however, the results remain controversial. This meta‐analysis was conducted to evaluate potential benefits of onion on lipid profile. Methods Up to 12 October 2020, PubMed, Cochrane Library, Web of Science, and Scopus were searched for randomized controlled trials evaluating the effects of onion on lipid profile. Mean differences (MD) and 95% confidence intervals (CI) were calculated. Meta‐analysis was conducted using the fixed‐effects model. Results Ten trials with 446 participants in total were included in the meta‐analysis. The pooled findings of 10 studies suggested that onion supplementation significantly improved high‐density lipoprotein cholesterol (HDL) (MD: 2.29 mg/dl; 95% CI: 0.87, 3.72; I 2 = 0%) and low‐density lipoprotein cholesterol (LDL) (MD: −6.64 mg/dl; 95% CI: −10.91, −2.36; I 2 = 32%),while onion supplementation did not significantly lower triglycerides (TG) (MD: −6.55 mg/dl; 95% CI: −15.64, 2.53; I 2 = 45%). Analysis of nine trials showed a significant reduction in total cholesterol (TC) (MD: −5.39 mg/dl; 95% CI: −10.68, −0.09; I 2 = 49%) in patients with onion supplementation compared to the control group. Conclusion In summary, supplementation of onion was beneficial to control dyslipidemia, including improving levels of HDL, LDL, and TC, but could not reduce TG level. The therapeutic benefits of onion for dyslipidemia need to be treated with caution considering that some of the results are not robust.

. Therefore, the treatment of dyslipidemia is one of the keys to prevent cardiovascular diseases (Yandrapalli et al., 2019).
Statins, the main lipid-lowering agents, are effective in improving blood lipids, but they are associated with many side effects that affect patient compliance (Karr, 2017). Furthermore, the reduction of drug efficacy is also a problem that cannot be ignored when statins are used for a long-term (Shekarchizadeh-Esfahani et al., 2020). Therefore, it is urgent to find a safe and effective supplementary treatment.
Alternatives based on natural products are of great interest because of their safety and potential effectiveness (Georgia-Eirini et al., 2019;Yarla et al., 2016). Onion (Allium Cepa L.) is a perennial plant, belonging to Family Amaryllidaceae, one of the most widely consumed vegetables in the world (Khajah et al., 2019;Li et al., 2020). Onions are rich in sulfur compounds and Flavonoids (Chiu et al., 2016), and have traditionally been used to treat asthma, coughs, high blood pressure, ulcer wounds, and other ailments (Khajah et al., 2019). In recent years, onion has been found to have many kinds of biological activities, including anti-inflammatory, antioxidant, anticancer, anti-diabetes, Immunoprotective, wound healing, anti-scar and, anti-obesity, and widely used in medicine (Khajah et al., 2019;Teshika et al., 2019). Recently, several studies have examined the effects of onion on blood lipids, but the results are contradictory. Lee et al. (2011) found that supplementation with 228 mg onion skin extract for 10 weeks significantly improved TC, LDL, and HDL levels in healthy subjects, but in Kim et al.'s study (2015), there was no significant difference in lipid profile after 12 weeks of administration of onion peel extract of 100 mg/day compared with placebo.
Therefore, it is necessary to summarize the evidence of the effects of onion on blood lipids. Unfortunately, there has been no systematic review or meta-analysis in this area. The purpose of this study was to comprehensively review randomized controlled trials to evaluate the effects of onion on lipid profile.

| Study selection
Trials meeting the following criteria were included: (a) Participants with dyslipidemia or without dyslipidemia; (b) Intervention were onion or products of onion; (c) The control group was placebo; (d) Outcomes assessed at least one of LDL, HDL, TG, or TC; (e) The

Database
Search term (establish to 12 October 2020) Number

| Data extraction
Characteristics of the included study were extracted, including first author, publication year, location, age, sample size, baseline body mass index (BMI), state of health, type of study, intervention, duration, outcomes. In addition, the mean changes of TG, TC, HDL, and LDL from baseline to the end of the study and their standard deviations (SDs) were also extracted. When data in the literature was not available, we contacted the corresponding author and attempt to obtain the data.

| Quality assessment
The whole process of literature search, screening, data extraction, and quality assessment was conducted by two authors (Tang and Tao) independently, and all inconsistencies were discussed with the third author (Zhang) and resolved. Cochrane Collaboration's tool was used to conduct quality assessments, including the following seven domains: (a) random sequence generation; (b) allocation concealment; (c) blinding of participants and personnel; (d) blinding of outcome assessment; (e) incomplete outcome data; (f) selective reporting; (g) other sources of bias. Each domain was classified as high bias risk, low bias risk, and unclear bias risk. Based on the domains mentioned, the overall quality of each study was assessed as good (more than two domains were assessed as low risk), fair (two domains were assessed as low risk), or weak (less than two domains were assessed as low risk).

| Statistical analysis
All concentration units were converted to mg/dl before the total effect is calculated. Effect size was defined as mean difference (MD) and 95% confidence interval (CI) (Higgins & Green, 2011. Available from www.cochr ane-handb ook.org). When no net change in the outcomes (TG, TC, LDL, and HDL) was provided, the difference between baseline and end is calculated as the effect size. The following formula was used to calculate standard deviations (SDs) of the mean changes: 5. When standard error (SE) was reported, the following formula is used to calculate SD: SD = SE × √n (n = number of F I G U R E . 1 PRISMA flow diagram of the literature retrieval process subjects). I 2 statistics were used to assess the size of heterogeneity. When I 2 > 50 was significant heterogeneity, and the randomeffect model was selected, whereas the fixed-effect model was selected (Higgins et al., 2003). We separately excluded each study to explore the robustness of our results. Begg's rank correlation and Egger's weighted regression were used to examine potential publication bias. Subgroup analysis was performed based on the type of participants (dyslipidemia or without dyslipidemia) and period of treatment (>10 weeks or ≤10 weeks). All statistical analyses were performed using Review Manager 5.3 (The Nordic Cochrane Centre, The Cochrane Collaboration 2014; Copenhagen, Denmark) and Stata 12.0 (Stata Corp.). Moreover, statistically significant is defined as p < .05.

| Identification and selection of studies
Of the 3,365 related literatures, 1,242 duplicates were excluded.
Next, 2,096 literatures that did not meet the inclusion criteria were excluded by reviewing the title and abstract. Then, we evaluated the remaining 27 articles, and finally, 10 articles were included ( Figure 1).

| Study characteristics
A total of 446 participants in the 10 included studies were randomly Characteristics of the included study are summarized in Table 2.

| Quality assessment
All of the included studies mentioned randomization, but only

| Subgroup analysis
Notably, the subgroup analysis showed that in the subgroup of subjects with dyslipidemia, onion intake appeared to show a greater In addition, we observed more increases in HDL concentrations (MD: 2.81 mg/dl; 95% CI: 1.08, 5.54; I 2 = 17%) in the subgroup with onion intake longer than 10 weeks. The results of subgroup analysis are summarized in Table 3.

| Publication bias
There was no evidence of publication bias for TG (Begg's = 1.000,

| D ISCUSS I ON
To our knowledge, this is the first meta-analysis to assess the impact of onion on blood lipids. Our results showed that onion could not reduce plasma TG levels, but onion administration did improve Onions are rich in flavonoids and dietary fiber, which epidemiology has shown can reduce the risk of cardiovascular disease (Hamauzu et al., 2011). However, the clinical evidence for the efficacy of onion on blood lipids is not uniform. Some studies have shown that onion consumption is effective in improving blood lipid levels in both patients with hyperlipidemia and healthy subjects (Louria et al., 1985;Vidyashankar et al., 2010). Chiu et al. (Chiu et al., 2016) found that red wine onion extract showed additional lipid-lowering effects compared with red wine after supplementing red wine onion extract at 250 ml/day for 10 weeks for hyperlipidemia patients. Lee et al. showed that with the intake of frozen onion powder, the subjects' plasma HDL level increased, levels of plasma TC, plasma LDL, and atherosclerosis index significantly decreased (Lee et al., 2008), which was similar to our questionnaire for participants showed that most patients with hyperlipidemia thought onion intake was good for their health (Lee et al., 2008). However, Arora's study (Arora & Arora, 1981) did not observe the benefit of onion supplementation on dyslipidemia.
Similarly, Ebrahimi-Mamaghani et al. found that eating raw red onion did not improve HDL and TG levels in patients with polycystic ovary syndrome (Ebrahimi-Mamaghani et al., 2014). Different characteristics of the subjects (healthy subjects, hyperlipidemia subjects), different onion products (onion skin extract, raw onion, onion juice), and Inconsistency in the duration of onion supplementation may lead to inconsistent results. As the results of the subgroup analysis showed, onion appeared to be more beneficial for subjects with dyslipidemia, and onion supplementation for longer than 10 weeks may be more beneficial for improving HDL.
Since eight studies used onion extract, one used red onion, and one used onion juice, subgroup analysis based on different onion products was not feasible.
Although a large number of studies have investigated the effects of onion on blood lipids, the mechanism of onion improving blood lipids is still unclear and may be related to the following aspects. On the one hand, onion can activate lecithin-cholesterol Acyltransferase by enhancing the action of insulin, so as to promote the conversion of LDL into HDL (Ige & Akhigbe, 2013). On the other hand, onion can promote the excretion of bile acids and inhibit the absorption of cholesterol to reduce plasma cholesterol (Guan et al., 2010). In addition, the anti-lipid effect of onion may be related to reducing lipid hydroperoxide and lipoperoxide concentrations (Campos et al., 2003;Guan et al., 2010).
Natural products are increasingly popular because they are believed to have few side effects (Teshika et al., 2019

| CON CLUS IONS
In summary, onion supplementation was beneficial to control dyslipidemia, including improving plasma levels of HDL, LDL, and TC, but did not reduce plasma TG level. The therapeutic benefits of onion for dyslipidemia need to be treated with caution considering that some of the results are not robust. It is necessary to investigate the effect of long-term onion supplementation on dyslipidemia through large-sample, high-quality studies. In addition, onion may be more effective in subjects with dyslipidemia, and future studies should focus on patients with hyperlipidemia to better assess the effect of onion on blood lipid.

ACK N OWLED G M ENTS
None declared.

CO N FLI C T S O F I NTE R E S T
The authors declare that they do not have any conflict of interest.

E TH I C A L A PPROVA L
This study does not involve any human or animal testing.

DATA AVA I L A B I L I T Y S TAT E M E N T
Data sharing is not applicable to this article as no new data were created or analyzed in this study.