Abstract
To investigate the effects of rapeseed oil on body composition, blood glucose and lipid metabolism in people with overweight and obesity compared to other cooking oils. We searched eight databases for randomized controlled studies (including randomized crossover trials). The risk of bias for the included studies was assessed using the Cochrane Risk of Bias 2.0 tool. The Grading of Recommendations Assessment Development and Evaluation (GRADE) criteria were used to evaluate the quality of the outcomes. The methodological quality of the included studies was assessed using the Physiotherapy Evidence Database (PEDro) scale. Sensitivity analysis was used to check the stability of the pooled results. Statistical analysis was carried out using Review Manager 5.3 software. As a result, fifteen randomized controlled studies (including six parallel studies and nine crossover studies) were included in this study. Compared to other edible oils, rapeseed oil significantly reduced low density lipoprotein cholesterol (LDL-C) (MD = −0.14 mmol/L, 95% CI: −0.21, −0.08, I2 = 0%, P < 0.0001), apolipoprotein B (ApoB) (MD = −0.03 g/L, 95% CI: −0.05, −0.01, I2 = 0%, P = 0.0003), ApoB/ApoA1 (MD = −0.02, 95% CI: −0.04, −0.00, I2 = 0%, P = 0.02) and insulin (MD = −12.45 pmol/L, 95% CI: −19.61, −5.29, I2 = 37%, P = 0.0007) levels, and increased fasting glucose (MD = 0.16 mmol/L, 95% CI: 0.05, 0.27, I2 = 27%, P = 0.003) levels. However, the differences in body weight and body composition between rapeseed oil and control oils were not significant. In a word, rapeseed oil is effective in reducing LDL-C, ApoB and ApoB/ApoA1 levels in people with overweight and obesity, which is helpful in preventing and reducing the risk of atherosclerosis. PROSPERO registration number: CRD42022333436.
Similar content being viewed by others
The effects of low-fat, high-carbohydrate diets vs. low-carbohydrate, high-fat diets on weight, blood pressure, serum liquids and blood glucose: a systematic review and meta-analysis
Article24 June 2021
Red meat consumption and risk factors for type 2 diabetes: a systematic review and meta-analysis of randomized controlled trials
ArticleOpen access05 May 2022
Dietary fats and cardiometabolic disease: mechanisms and effects on risk factors and outcomes
Article16 May 2019
Introduction
Over the past 50 years, there has been a significant increase in obesity rates worldwide. The GBD (Global Burden of Disease) Obesity Collaborators estimated that 603.7 million adults plagued by obesity, with the obesity prevalence doubling in 73 countries between 1980 and 2015 and continuing to rise in most other countries [1]. A high-fat diet usually leads to the development of obesity [2]. With the improvement of living standards, there have been important changes in people’s eating habits, such as unhealthy diets with high sugar, oil, and fat contents, and the resulting potential for obesity problems [3,4,5]. The prevalence of obesity is increasing in younger populations, with the prevalence increasing not only in older people [6,7,8] but also in children [3, 9,10,11].
Obesity is often thought of as an excess accumulation of body fat. According to WHO standards, the diagnostic criteria for overweight and obesity in adults are as follows: overweight: 25 kg/m2≤body mass index (BMI) < 30 kg/m2 and obesity: BMI ≥ 30 kg/m2 [12]. However, these criteria do not well reflect the relationship between BMI and overweight or obesity in Asian populations (e.g., China) due to the differences in body composition of different ethnic populations. Therefore, for the Chinese population, a threshold of 24 kg/m2 for overweight and 28 kg/m2 for obesity is considered to be more appropriate [13]. Obesity can cause many negative effects on human health. There is a strong link between obesity and type 2 diabetes and cardiovascular disease [14]. In addition, obesity is considered to be an important cause of metabolic syndrome [15, 16], and can lead to hyperglycemia [17, 18], hyperlipidemia [19], and hypertension [20, 21]. Therefore, the prevention and treatment of obesity is a matter of urgency.
As mentioned earlier, unhealthy eating patterns are a potential source of obesity. Therefore, dietary interventions may be a good way to prevent and treat obesity [22, 23]. Rapeseed oil (canola oil) contains very few saturated fatty acids (SFAs) and is rich in monounsaturated fatty acids (MUFAs) and polyunsaturated fatty acids (PUFAs), including 61% oleic acid, 21% linoleic acid, and 11% alpha-linolenic acid (ALA), which are beneficial to the human body [24]. Studies have shown that, compared to SFAs, MUFAs can be involved in the regulation of cardiovascular health by regulating plasma lipids and lipoproteins, susceptibility to low-density lipoprotein (LDL) oxidation, and insulin sensitivity [25,26,27]. A MUFA-rich diet can reduce total cholesterol (TC) and LDL levels in people with moderately obese [28], as well as the body weight and fat mass [29]. The mechanism by which MUFA reduces obesity-related index may be through increased oxidation rate and energy expenditure [30]. The increase in fatty acid oxidation capacity may be due to the activation of peroxisome proliferator-activated receptor δ (PPAR-δ) by the consumption of MUFA [31, 32]. Furthermore, as a derivative of oleic acid, oleoylethanolamide can effectively activate PPAR-α and lead to lipolysis [33]. Therefore, a long-term high MUFA diet may lead to changes in body composition. Given these benefits of rapeseed oil, it is increasingly being used in the management and treatment of overweight and obesity [34, 35]. Recently, a growing number of scholars have explored the effects of rapeseed oil on body composition and oxidative metabolism in patients with obesity and overweight [30, 34, 36]. However, different studies have come to different conclusions, and some even show opposite results. A randomized double-blind crossover trial that included 44 patients with overweight and obesity showed that rapeseed oil did not have a greater beneficial effect than amaranth oil on atherosclerosis markers [37]. However, another randomized crossover trial with a larger sample size and longer intervention time showed that conventional and high oleic rapeseed oils were effective in improving lipid and lipoprotein parameters [36]. The moderate intake of rapeseed oil is an effective strategy to reduce the risk of atherosclerotic cardiovascular disease [36]. Therefore, the beneficial effects of rapeseed oil in people with overweight and obesity remain controversial. In addition, several of the trials were small sample size studies. Hence, a meta-analysis and systematic review of these studies is needed to further investigate the effects of rapeseed oil on body composition and metabolism in patients with obesity and overweight.
Slider with three content items shown per slide. Use the Previous and Next buttons to navigate the slides or the slide controller buttons at the end to navigate through each slide
Abstract
To investigate the effects of rapeseed oil on body composition, blood glucose and lipid metabolism in people with overweight and obesity compared to other cooking oils. We searched eight databases for randomized controlled studies (including randomized crossover trials). The risk of bias for the included studies was assessed using the Cochrane Risk of Bias 2.0 tool. The Grading of Recommendations Assessment Development and Evaluation (GRADE) criteria were used to evaluate the quality of the outcomes. The methodological quality of the included studies was assessed using the Physiotherapy Evidence Database (PEDro) scale. Sensitivity analysis was used to check the stability of the pooled results. Statistical analysis was carried out using Review Manager 5.3 software. As a result, fifteen randomized controlled studies (including six parallel studies and nine crossover studies) were included in this study. Compared to other edible oils, rapeseed oil significantly reduced low density lipoprotein cholesterol (LDL-C) (MD = −0.14 mmol/L, 95% CI: −0.21, −0.08, I2 = 0%, P < 0.0001), apolipoprotein B (ApoB) (MD = −0.03 g/L, 95% CI: −0.05, −0.01, I2 = 0%, P = 0.0003), ApoB/ApoA1 (MD = −0.02, 95% CI: −0.04, −0.00, I2 = 0%, P = 0.02) and insulin (MD = −12.45 pmol/L, 95% CI: −19.61, −5.29, I2 = 37%, P = 0.0007) levels, and increased fasting glucose (MD = 0.16 mmol/L, 95% CI: 0.05, 0.27, I2 = 27%, P = 0.003) levels. However, the differences in body weight and body composition between rapeseed oil and control oils were not significant. In a word, rapeseed oil is effective in reducing LDL-C, ApoB and ApoB/ApoA1 levels in people with overweight and obesity, which is helpful in preventing and reducing the risk of atherosclerosis. PROSPERO registration number: CRD42022333436.
Similar content being viewed by others
The effects of low-fat, high-carbohydrate diets vs. low-carbohydrate, high-fat diets on weight, blood pressure, serum liquids and blood glucose: a systematic review and meta-analysis
Article24 June 2021
Red meat consumption and risk factors for type 2 diabetes: a systematic review and meta-analysis of randomized controlled trials
ArticleOpen access05 May 2022
Dietary fats and cardiometabolic disease: mechanisms and effects on risk factors and outcomes
Article16 May 2019
Introduction
Over the past 50 years, there has been a significant increase in obesity rates worldwide. The GBD (Global Burden of Disease) Obesity Collaborators estimated that 603.7 million adults plagued by obesity, with the obesity prevalence doubling in 73 countries between 1980 and 2015 and continuing to rise in most other countries [1]. A high-fat diet usually leads to the development of obesity [2]. With the improvement of living standards, there have been important changes in people’s eating habits, such as unhealthy diets with high sugar, oil, and fat contents, and the resulting potential for obesity problems [3,4,5]. The prevalence of obesity is increasing in younger populations, with the prevalence increasing not only in older people [6,7,8] but also in children [3, 9,10,11].
Obesity is often thought of as an excess accumulation of body fat. According to WHO standards, the diagnostic criteria for overweight and obesity in adults are as follows: overweight: 25 kg/m2≤body mass index (BMI) < 30 kg/m2 and obesity: BMI ≥ 30 kg/m2 [12]. However, these criteria do not well reflect the relationship between BMI and overweight or obesity in Asian populations (e.g., China) due to the differences in body composition of different ethnic populations. Therefore, for the Chinese population, a threshold of 24 kg/m2 for overweight and 28 kg/m2 for obesity is considered to be more appropriate [13]. Obesity can cause many negative effects on human health. There is a strong link between obesity and type 2 diabetes and cardiovascular disease [14]. In addition, obesity is considered to be an important cause of metabolic syndrome [15, 16], and can lead to hyperglycemia [17, 18], hyperlipidemia [19], and hypertension [20, 21]. Therefore, the prevention and treatment of obesity is a matter of urgency.
As mentioned earlier, unhealthy eating patterns are a potential source of obesity. Therefore, dietary interventions may be a good way to prevent and treat obesity [22, 23]. Rapeseed oil (canola oil) contains very few saturated fatty acids (SFAs) and is rich in monounsaturated fatty acids (MUFAs) and polyunsaturated fatty acids (PUFAs), including 61% oleic acid, 21% linoleic acid, and 11% alpha-linolenic acid (ALA), which are beneficial to the human body [24]. Studies have shown that, compared to SFAs, MUFAs can be involved in the regulation of cardiovascular health by regulating plasma lipids and lipoproteins, susceptibility to low-density lipoprotein (LDL) oxidation, and insulin sensitivity [25,26,27]. A MUFA-rich diet can reduce total cholesterol (TC) and LDL levels in people with moderately obese [28], as well as the body weight and fat mass [29]. The mechanism by which MUFA reduces obesity-related index may be through increased oxidation rate and energy expenditure [30]. The increase in fatty acid oxidation capacity may be due to the activation of peroxisome proliferator-activated receptor δ (PPAR-δ) by the consumption of MUFA [31, 32]. Furthermore, as a derivative of oleic acid, oleoylethanolamide can effectively activate PPAR-α and lead to lipolysis [33]. Therefore, a long-term high MUFA diet may lead to changes in body composition. Given these benefits of rapeseed oil, it is increasingly being used in the management and treatment of overweight and obesity [34, 35]. Recently, a growing number of scholars have explored the effects of rapeseed oil on body composition and oxidative metabolism in patients with obesity and overweight [30, 34, 36]. However, different studies have come to different conclusions, and some even show opposite results. A randomized double-blind crossover trial that included 44 patients with overweight and obesity showed that rapeseed oil did not have a greater beneficial effect than amaranth oil on atherosclerosis markers [37]. However, another randomized crossover trial with a larger sample size and longer intervention time showed that conventional and high oleic rapeseed oils were effective in improving lipid and lipoprotein parameters [36]. The moderate intake of rapeseed oil is an effective strategy to reduce the risk of atherosclerotic cardiovascular disease [36]. Therefore, the beneficial effects of rapeseed oil in people with overweight and obesity remain controversial. In addition, several of the trials were small sample size studies. Hence, a meta-analysis and systematic review of these studies is needed to further investigate the effects of rapeseed oil on body composition and metabolism in patients with obesity and overweight.
Methods
This study was performed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [38]. This article has been registered on PROSPERO under registration number CRD42022333436.
Search strategies
We searched the published literature in PubMed, Web of Science, Embase, Cochrane Library, Scopus, SinoMed, CNKI, and Wanfang Database without imposing publication date or language restrictions. A systematic search was performed for eligible studies published through April 21, 2022. The keywords were (rapeseed oil OR oil, rapeseed OR low erucic acid rapeseed oil OR LEAR oil OR oil, LEAR OR canola oil OR oil, canola OR colza oil OR rap oil) AND (obesity OR obese OR adiposis OR overweight) AND (randomized controlled OR RCT OR controlled trial OR clinical trial). The PubMed search strategy is shown in Table S1. In addition, to prevent the omission of relevant literature, references were searched for inclusion in the study.
Study selection
Two authors (YL, HY) independently selected the literature. If any differences existed, a meeting was held to resolve them. The inclusion criteria for the studies were based on the PICOS (patients, intervention, comparison, outcomes, and study design) principle [39], as shown below:
Patients (P): People with obesity or overweight (with or without metabolic risk). There was no age limit for participants, nor was there a limit on sex or environment (such as hospitals, communities, or nursing homes).
Intervention (I): The intervention was a rapeseed oil diet.
Comparison (C): Other oils, such as amaranth oil, olive oil and so on.
Outcomes (O): Primary outcomes: body weight (BW) and body composition (e.g., waist circumference (WC), body fat % (BF%), BMI); secondary outcomes: blood glucose and lipid metabolism markers (e.g., triglyceride (TG), TC, glucose, insulin).
Study design (S): Randomized parallel controlled trials and randomized cross-controlled trials were included in this study.
Studies were excluded if they met the following criteria: (1) the participants were not patients with obesity or overweight; (2) the intervention did not include a rapeseed oil diet; (3) the comparisons were not performed according to the intervention type; (4) the outcomes did not include BW, body composition, blood glucose or lipid metabolism; and (5) the study type was not a randomized controlled trial (such as a conference paper, a protocol, a review, or a case report).
Data extraction
Two authors (YLW, QZ) independently extracted the data from the selected studies into a Microsoft Excel spreadsheet and then summarized the data in a table. Any disagreement was resolved in a consensus meeting.
The data extracted from each study included the following: (1) the first author and year of publication; (2) study design; (3) the countries of the participants; (4) the gender and age of the participants; (5) the diagnostic criteria of obesity and overweight; (6) the groups and sample size; (7) the time points of assessment; (8) the total intervention time; (9) the outcome indicators (including primary outcomes and secondary outcomes); and (10) Physiotherapy Evidence Database (PEDro) scores.
Quality assessment
Two authors (JHZ, HH) independently completed an assessment of the methodological quality of each included study using the PEDro scale. If the score was inconsistent between the two authors, a third author (YL) was consulted to determine the final score. The PEDro scale has 11 items: (1) eligibility criteria and source; (2) random allocation; (3) concealed allocation; (4) baseline comparability; (5) participant blinding; (6) therapist blinding; (7) assessor blinding; (8) adequate follow-up (>85%); (9) intention-to-treat analysis; (10) between-group statistical comparisons; and (11) point and variability measurements. A total PEDro score is achieved by adding the ratings of items (2) to (11) for a combined total score between 0 and 10. Scores of less than 4 are considered poor, scores of 4 to 5 are considered fair, scores of 6 to 8 are considered good and scores of 9 to 10 are considered excellent [40].
Furthermore, we assessed the quality of the outcomes according to the Grading of Recommendations Assessment Development and Evaluation (GRADE) criteria, which are graded as very low, low, moderate and high [41]. For randomized trials, there are five indicators to assess the quality of the outcomes: risk of bias, inconsistency, imprecision, indirectness, and publication bias.
Risk of Bias
Two authors (YL, HY) assessed the risk of bias of the included studies using the Cochrane Risk of Bias (RoB) 2.0 Tool [42]. If there was a disagreement, a third author (YLW) was involved to reach a consensus. RoB 2.0 includes six domains, which are (1) randomization process, (2) deviations from intended interventions, (3) missing outcome data, (4) measurement of the outcome, (5) selection of the reported result, and (6) overall bias. Each domain was classified as “low risk”, “some concerns”, “high risk”. It is worth noting that the overall bias is judged according to the degree of bias in the previous five domains. The overall risk of bias is judged as follows: (1) Low risk of bias: The study is judged to be at low risk of bias for all domains for this result. (2) Some concerns: The study is judged to be at some concerns in at least one domain for this result. (3) High risk of bias: The study is judged to be at high risk of bias in at least one domain for this result, or, the study is judged to have some concerns for multiple domains in a way that substantially lowers confidence in the result. In addition, if more than 8 studies were included, funnel plots were used to assess publication bias.
Data synthesis and statistical analysis
Statistical analysis was performed using Review Manager 5.3 (The Nordic Cochrane Centre, Copenhagen, Denmark). The pre-extracted mean value, standard deviation and sample size were input into the statistical software. If the standard error (SE) of the data was given in the original text, the standard deviation (SD) was calculated according to Equation ①. If a study gave the SDs of baseline, endpoint, and change for partial data, according to the 16.1.3.2 item of Review Manager 5.3 handbook, the correlation coefficient R-value of the study can be calculated according to Equation ②. Then, based on this formula, the data for the missing endpoint values in the study could be calculated. If the R-value could not be calculated based on the data in the study, R = 0.8 was estimated. If a study provided data in median (interquartile range) form, the median was treated as the mean value and the SD was calculated according to Equation ③. If a study presented results in the form of graphs (e.g., line graphs, bar graphs), the WebPlotDigitizer 4.5 software was used to extract the raw data from the graphs (https://apps.automeris.io/wpd/index.zh_CN.html). In the formulas, n represents the sample size; SD(b), SD(e), and SD(c) represent the SD of baseline, endpoint, and change, respectively; Q1 represents the lower quartile and Q3 represents the upper quartile.
$$SD=SE\sqrt{n}$$
(1)
$$R=\frac{SD(b)2+SD(e)2-SD(c)2}{2\times SD(b)\times SD(e)}$$
(2)
$$SD=(Q3-Q1)\div1.35$$
(3)
In our meta-analyses, for the same continuous outcomes, if the units were the same, we used mean difference (MD) with 95% confidence intervals (95% CIs) to assess the effect size. If the units were different, they were converted to the same units. If the units could not be converted, standardized mean difference (SMD) was used to estimate the effect size. In all analyses, I2 statistics were used to analyze heterogeneity between studies. If the P value of the heterogeneity test (I2 statistic) was less than 0.05, the random effects model was used; otherwise, a fixed-effects model was used. For small sample studies and highly heterogeneous outcomes, subgroup analysis was used to look for sources of the heterogeneity, and sensitivity analysis was used to test the stability of the results.
Since this study included both parallel and crossover trials, when suitable data from crossover trials were available for meta-analysis, we used the method recommended by Elbourne et al. [43]. Where possible, we used data from paired analysis results for meta-analysis (i.e., information from participants’ own pre- and post-intervention comparisons, and appropriate SEs to estimate treatment effects). If this was not possible, we combined the data from the first stage (i.e., crossed over the previous data) as if they were derived from a parallel study design. When paired data or data from the first stage were not available, we treated the data from the crossover trials as if they were from parallel trials and then sequentially excluded the crossover trials for sensitivity analysis to assess the stability of the results.
Results
Study selection
For Chinese databases, we searched both Chinese and English literature. After a systematic search of eight databases, we identified a total of 540 articles, including 40 in PubMed, 14 in Embase, 33 in Web of Science, 75 in the Cochrane Library, 343 in Scopus, 35 in SinoMed, 0 in CNKI, and 0 in Wanfang Database. In addition, 6 papers were obtained from other sources. After deleting duplicate studies, 381 studies remained. After reading the titles and abstracts and excluding studies that did not meet the inclusion criteria, 21 papers remained. We then read the full text of these 21 papers to further eliminate studies that did not meet the requirements. As a result, 6 papers were excluded. The list of study exclusions and the reasons for exclusion are shown in Table S2. As a result, fifteen studies were included for qualitative analysis, including six randomized controlled trials [44,45,46,47,48,49] and nine randomized crossover controlled trials [30, 34, 36, 37, 50,51,52,53,54]. For the randomized crossover trials, we treated the final data from the crossover trials as coming from parallel trials because neither the paired data nor the first stage data were available. Hence, to minimize the effect of multiple phase trials on the results, we only extracted data from trials with less than or equal to 3-period crossover for the meta-analysis. Only qualitative descriptions were given for crossover studies with greater than 3-period crossover [34, 50, 51]. Therefore, we ultimately included twelve studies in the meta-analysis [30, 36, 37, 44,45,46,47,48,49, 52,53,54]. (Fig. 1)
Study characteristics
Among the fifteen included studies, the subjects in three studies were from Germany [44, 46, 47], four studies were from Poland [37, 45, 48, 54], one study was from the United Kingdom [49], two studies were from Iran [30, 52], one study was from Canada [36], one study was from USA [53], and those in three studies were from Canada and USA [34, 50, 51]. Twelve studies involved both men and women [30, 34, 36, 37, 44, 45, 48,49,50,51,52, 54], and the other three studies only included men to exclude possible confounding effects of estrogen on the results [46, 47, 53]. The duration of the intervention was 3 weeks in four studies [37, 45, 48, 54], 4 weeks in four studies [34, 47, 50, 51], 6 weeks in two studies [36, 53], 8 weeks in one study [46], 9 weeks in two studies [30, 52], 12 weeks in one study [49], and 6 months in one study [44]. Three studies compared rapeseed oil with olive oil [44, 46, 47], four studies compared rapeseed oil with amaranth oil [37, 45, 48, 54], one study compared rapeseed oil with sunflower oil [49], two studies compared rapeseed oil with sesame oil [30, 52], one study compared high oleic canola oil plus medium chain triglycerides with olive oil [53], one study had two types of experimental oils: canola oil and high oleic canola oil [36], and the other three studies were 5-period randomized crossover studies involving five experimental oils [34, 50, 51]. For diagnostic criteria of overweight and obesity, three studies were based on WC [36, 44, 50], nine studies on BMI [30, 37, 45,46,47,48, 52,53,54], and three studies on WC combined with BMI [34, 49, 51]. (Table 1)Table 1 Characteristics of the included randomized controlled trials.
The funding status and conflicts of interest of the included studies are as follows. Ten studies received funding [30, 37, 44, 46, 47, 49, 50, 52,53,54], five studies had potential conflicts of interest [37, 46, 49, 50, 54], and three studies did not mention funding or conflicts of interest [34, 36, 51].
Dietary intervention protocol
One study required participants to consume 30 grams of the study oil and 20 grams of margarine per day [44]. In addition, to compensate for the lower linoleic acid content of olive oil compared to that of rapeseed oil, patients in the olive oil group consumed sunflower oil once a week [44]. One study required participants to consume 20 ml of rapeseed oil or amaranth oil per day and a diet of 1800 kcal [45]. Furthermore, these oils were given to patients at the same time of day (mid-morning) to avoid the effects of circadian rhythms [45]. Subjects in two studies had to consume 50 g of rapeseed or olive oil per day [46, 47]. Subjects in one study added 20 ml of rapeseed oil or amaranth oil to their daily diet and received aerobic exercise and physical training guided by a physiotherapist [48]. Participants in one study were asked to consume 20 ml of the experimental oil, uncooked [49]. For the randomized crossover study, one study was a three-phase crossover trial with a 6-week intervention and a 4-week washout period with an energy intake of 3,000 kcal [36]. A two-phase crossover study (two articles) for 3 weeks with 3-week washout period required participants to consume 20 ml of the experimental oil daily [37, 54]. Three studies were five-phase crossover trials containing five intervention oils, each phase of intervention for 4 weeks, with washout periods of 2 to 4 weeks [34, 50, 51]. Two articles (the same study) asked participants to intake 30 g of the experimental oil daily for 9 weeks with 4-week washout period [30, 52]. One study consisted of two separate phases, each with a 6-week intervention and a washout period of 4 to 8 weeks [53]. (Table S3) The fatty acid compositions of the treated oils of the studies by Moszak et al. 2020 [48], Kruse et al. 2020 [46] and Jones et al. 2014 [50] are shown in Table S4.
Quality of the included studies
The PEDro scores for each study are shown in Table 1, and the specific scoring details are shown in Table S5. Four studies scored 9, and the quality of the studies was considered to be excellent [30, 37, 52, 54]. Eight studies scored 6 to 8, and the quality was considered good [34, 36, 45, 47,48,49,50,51]. Three studies scored 5, and the quality was considered fair [44, 46, 53]. All fifteen studies had random allocation, baseline comparability, between-group statistical comparisons, and point and variability measurements. Eight studies used concealed allocation [30, 34, 37, 49,50,51,52, 54]. Seven studies were double-blind [34, 36, 37, 48, 50, 51, 54], two studies were triple-blind [30, 52], and two studies were single-blind [49, 53]. Ten studies had a sufficient number of follow-ups (>85%) [30, 36, 37, 44,45,46,47,48, 52, 54]. All subjects in the four studies completed the experiments according to the established study protocol [37, 45, 47, 54].
Risk of bias of the included studies
The risk of bias assessment for the fifteen studies is shown in Fig. 2 according to the Cochrane RoB 2.0 Tool. Because of the different designs of parallel and crossover trials, we used the RoB 2.0 for parallel trial and the RoB 2.0 for crossover trial to assessment risk, respectively. For the parallel trials, there was a high risk of randomization process in five studies [44,45,46,47,48], a high risk of missing data in one study [48], and some concerns about deviations from intended interventions in all six studies [44,45,46,47,48,49]. Thus, the overall bias was high risk for 5 parallel studies [44,45,46,47,48] and some concern for 1 parallel study [49]. For the crossover trials, there was a high risk for the randomization process in one study [53], some concern for the randomization process in six studies [34, 36, 37, 50, 51, 54], and some concern for missing data in four studies [34, 36, 50, 51]. Therefore, the overall bias was high risk for 1 crossover study [53], some concern for 6 crossover studies [34, 36, 37, 50, 51, 54], and low risk for 2 crossover studies [30, 52].
Effects of rapeseed oil on primary outcomes
Compared to other cooking oils, ten studies investigated the effects of rapeseed oil on BW [30, 36, 44,45,46,47,48,49, 53, 54], eight studies investigated BMI [30, 44,45,46,47,48,49, 54], five studies assessed WC [30, 44, 48, 49, 54], five studies measured BF% [30, 44, 46,47,48], six studies investigated the waist-hip ratio (WHR) [30, 45,46,47,48, 54], and two studies measured fat free mass (FFM) [45, 48].
The forest plot results showed that rapeseed oil did not significantly improve BW, BMI, WC, the WHR or FFM in individuals with overweight or obesity compared to other cooking oils (P > 0.05). For BF%, the P value was at a critical value and had a tendency to favor the control group (MD = 0.80%, 95% CI: 0.02, 1.58, I2 = 41%, P = 0.05) (Fig. 3). Similarly, Jones et al. 2014 showed that the difference in BW between all groups after the dietary intervention was not significant [50]. However, a study by Liu et al. 2016 showed that the rapeseed oil (P = 0.007) and high oleic rapeseed oil (P = 0.02) groups significantly reduced BW compared to the Flax/Saff (flax/safflower oil) group [34].
Effects of rapeseed oil on secondary outcomes
Compared to other cooking oils, nine studies investigated the effect of rapeseed oil on TC, TG, low-density lipoprotein cholesterol (LDL-C) and high-density lipoprotein cholesterol (HDL-C) levels [36, 44, 46,47,48,49, 52,53,54]. Five studies evaluated the effect of rapeseed oil on homeostatic model assessment of insulin resistance (HOMA-IR) scores [37, 46,47,48, 52], seven studies examined fasting blood glucose [37, 44, 46,47,48,49, 52], five studies measured serum insulin levels [44, 46,47,48, 52], and four studies assessed apolipoprotein A1 (ApoA1), apolipoprotein B (ApoB) levels and ApoB/ApoA1 ratio [36, 37, 44, 52].
The results of the forest plot showed that rapeseed oil significantly reduced LDL-C (MD = −0.14 mmol/L, 95% CI: −0.21, −0.08, I2 = 0%, P < 0.0001), insulin (MD = −2.45 pmol/L, 95% CI: −19.61, −5.29, I2 = 37%, P = 0.0007), ApoB (MD = −0.03 g/L, 95% CI: −0.05, −0.01, I2 = 0%, P = 0.0003) and ApoB/ApoA1 ratio (MD = −0.02, 95% CI: −0.04, −0.00, I2 = 0%, P = 0.02) compared to other oils. However, for blood glucose, the results showed in favor of the control group (MD = 0.16 mmol/L, 95% CI: 0.05, 0.27, I2 = 27%, P = 0.003). Besides these, the differences between the experimental and control groups were not significant for TC, TG, HDL-C, HOMA-IR and Apo A1 (P > 0.05) (Fig. 4). Jones et al. 2014 showed that the end of CanolaDHA (high–oleic acid canola oil with DHA) feeding resulted in higher endpoint values for TC, LDL-C, HDL-C, ApoA1, and ApoB, but resulted in lower endpoint values for TG [50]. The canola oil group was second only to the CanolaDHA group in most endpoint values, both increased and decreased. The study of Liu et al. 2018 showed that after 4 weeks, all five diets decreased TC, LDL-C, and TG concentrations (P < 0.05 for all) from baseline [51].
Quality of outcome indicators
Using the GRADE profiler 3.6.1, the level of evidence for outcome quality was categorized into four levels: very low, low, moderate and high. With the exception of ApoA1, ApoB and ApoB/ApoA1 ratio, the remaining outcomes were at serious risk of bias, as most studies were at high risk of overall bias. TC had a serious inconsistency because of the relatively high heterogeneity (I2 = 57%). WHR, FFM and HOMA-IR had serious imprecision because the sample size was less than 400 individuals. Glucose was found to have the potential for publication bias because of asymmetry on both sides of the funnel plot. BF% and insulin was found to have the potential for publication bias because one study had a small sample size and received funding, but the weight in the meta-analysis was large [47]. None of the outcomes had serious indirectness. As a consequence, for primary outcomes, the quality of evidence for BW, BMI and WC was moderate, but that for BF%, WHR and FFM was low. Regarding the secondary outcomes, the quality of evidence for ApoA1, ApoB and ApoB/ApoA1 was high, and that for LDL-C, HDL-C and TG values was moderate, but that for insulin, glucose and HOMA-IR scores was low (Table 2).Table 2 GRADE evidence profile for primary outcomes and secondary outcomes among trials included in the systematic review.
Subgroup analysis
For outcomes with high heterogeneity (I2 > 50%), we performed subgroup analysis to find the source of heterogeneity and to explore the effect of different subgroups on the outcome. We noted that among all the outcomes, only TC had a high heterogeneity (I2 = 57%), and the remaining outcomes were relatively homogenous. In other words, the homogeneity between the parallel and crossover trials is better.
We established subgroups based on the different types of control oils. For CT, subgroup analysis showed that rapeseed oil had a tendency to lower TC compared to olive oil but not significantly (MD = −0.19 mmol/L, 95% CI: −0.38, 0.01, I2 = 0%, P = 0.06); compared to blended oil, rapeseed oil significantly lowered TC (MD = −0.18 mmol/L, 95% CI: −0.26, −0.10, I2 = 0%, P < 0.00001); however, compared to sunflower oil, it increased TC levels (MD = 0.60 mmol/L, 95% CI: 0.27, 0.93, I2 = 0%, P = 0.0004). Therefore, this heterogeneity mainly originated between subgroups, i.e., the types of control oils (Fig. S1).
Sensitivity analysis
To test the stability of the results of the meta-analysis, a sensitivity analysis of the outcome indicators was carried out. For BMI, when the study by Kruse et al. [47] was removed, the heterogeneity became 0%, and the difference in BMI between the experimental and control groups tended to be significant (MD = −0.70 kg/m2, 95% CI: −1.44, 0.05, I2 = 0%, P = 0.07). For WC, when the study of Baxheinrich et al. [44] was removed, rapeseed oil showed a trend of increasing WC, and P value was at critical value (MD = 2.19 cm, 95% CI: −0.01, 4.39, I2 = 0%, P = 0.05), compared to the control oil. For BF%, the result was meaningless (P > 0.05) when the study of Kruse et al. [47] was taken out. However, when any of the other studies were excluded, the results showed a BF% increase in rapeseed oil group and the results of the meta-analysis favored the control group. For TC, when the study of Nicol et al. [49] was excluded, rapeseed oil significantly reduced TC levels compared to the control oil (MD = −0.16 mmol/L, 95% CI: −0.23, −0.09, I2 = 0%, P < 0.00001), and the heterogeneity was reduced to 0%. For LDL-C, ApoB and ApoB/ApoA1, when the study of Bowen et al. 2019 [36] was removed, the difference between the experimental and control groups was not significant (P > 0.05). For glucose and insulin, when the study of Kruse et al. [47] was excluded, the difference in glucose and insulin levels between the experimental and control groups was not significant (P > 0.05) (Fig. S2).
Publication Bias
If the number of studies measuring the same outcome indicator was greater than or equal to 8, funnel plot analysis was performed to detect the presence of publication bias. The funnel plot results showed no significant publication bias for BW, BMI, TC, TG, HDL-C and LDL-C. However, for blood glucose, the funnel plot showed a significant publication bias. Besides, BF% and insulin was found to have the potential for publication bias because one study had a small sample size and received funding, but the weight in the meta-analysis was large [47] (Fig. 5).
Discussion
This systematic review and meta-analysis of randomized controlled trials revealed that, compared to the control oils, rapeseed oil intake significantly decreased LDL-C, insulin, ApoB and ApoB/ApoA1 ratio, and increased blood glucose level in people with overweight and obesity. For body composition, TC, TG, HDL-C, ApoA1 and HOMA-IR, the differences between rapeseed oil and control oil were not significant. The results of the subgroup analysis showed that, compared to olive oil, rapeseed oil had a tendency to decrease TC level, and compared to blended oil, it significantly reduced the TC level. However, rapeseed oil increased TC level compared to sunflower oil.
Rapeseed oil is rich in PUFAs and MUFAs [24]. ALA is one of the main PUFAs of rapeseed oil, which is an essential amino acid that can be metabolized to eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) [55]. Inflammation is the basis of many chronic diseases, including coronary heart disease, diabetes, arthritis, cancer and many others, and EPA and DHA have the most potent anti-inflammatory effects [55]. In addition, the beneficial effects of rapeseed oil may be derived from the high content of MUFAs, the high ratio of unsaturated to saturated fatty acids (15:1), and the appropriate ratio of omega-6 to omega-3 (2:1) [56]. In clinical dietary interventions, an appropriate reduction in omega-6 intake and an appropriate increase in omega-3 intake, i.e., a lower ratio of omega-6/omega-3, is thought to reduce the risk of many chronic diseases [55].
Being overweight and obese has long been a major concern. Regarding the diagnostic criteria for overweight and obesity, there are slight differences between Western countries and China, as mentioned in the Introduction section. WC measurement is considered to be a simple, inexpensive method for diagnosing abdominal obesity [57]. Similarly, there are ethnic differences in the diagnosis of abdominal obesity. For Europeans, the diagnosis of abdominal obesity is as follows: a WC ≥ 94 cm for men and ≥80 cm for women [58]. For the Chinese population, the diagnosis of abdominal obesity is as follows: WC ≥ 90 cm for men and ≥80 cm for women [58].
There are various methods that people have used to try to prevent and improve overweight and obesity, such as resistance training [59, 60], physical activity [61, 62], electrical stimulation [63], and dietary interventions [64]. As rapeseed oil has shown beneficial effects in humans, many scholars have used it for the dietary management of overweight and obesity. Our study also further summarizes the evidence on the effects of rapeseed oil on body weight, body composition and glucose lipid metabolism in people with overweight and obesity compared to other edible oils.
To our knowledge, this study is the first systematic review and meta-analysis of randomized controlled trials on the effects of rapeseed oil on body composition and glucose lipid metabolism in patients with obesity and overweight. Our results show that rapeseed oil does not produce more beneficial effects on body weight and body composition than other oils. This is similar to the findings of Raeisi-Dehkordi et al., who found that rapeseed oil had no significant effect on body composition other than reducing body weight [56]. In addition, our results showed that rapeseed oil significantly reduced LDL-C, insulin, ApoB and ApoB/ApoA1 compared to other edible oils. This has some similarities to the results of a systematic review and meta-analysis that did not restrict recruitment to populations with overweight or obesity. It found that rapeseed oil significantly reduced TC, LDL-C, ApoB and ApoB/ApoA1 compared to other oils [65]. All these results may suggest that rapeseed oil does have a lowering effect on LDL-C, ApoB and ApoB/ApoA1 levels, which has an important role in preventing and reducing the risk of developing atherosclerosis. Therefore, we recommend adding rapeseed oil to the daily diet in moderation for people with overweight or obesity to prevent and reduce the risk of developing atherosclerosis. Furthermore, our study could not see that olive oil was more effective than rapeseed oil in improving body composition and reducing cardiovascular disease, which is similar to the findings of Raeisi-Dehkordi et al. [56] and Amiri et al. [65]. For the Chinese population, the market price of olive oil is much higher than the price of rapeseed oil. Due to the advertising effect, an increasing number of Chinese people are inclined to buy olive oil because of its beneficial effects on health. However, our studies have shown that the effects of olive oil on body composition and blood glucose and lipid metabolism in people with overweight or obesity do not differ significantly from those of rapeseed oil. Therefore, people with overweight or obesity who expect to prevent and improve their cardiovascular health by consuming olive oil should be aware that olive oil does not necessarily provide more beneficial effects than rapeseed oil. Moreover, olive oil is more expensive, whereas rapeseed oil is more affordable for the general population.
Strengths and limitations
Strength: The RoB 2.0 tool was used in this study to assess the risk of bias in the included studies, and compared to RoB 1.0, it has good specificity for randomized trials with parallel, crossover, and whole-group designs, respectively.
Limitations: First, the duration of the intervention was short (3 ~ 9 weeks) in most studies, with only one study having an intervention lasting six months. Second, for parallel trials, the overall bias for most studies was high risk, because the randomization process and the deviation from intended intervention domains were high risk of bias. Third, it is important to note that the interpretation of BF%, TC, LDL-C, ApoB, ApoB/ApoA1, glucose and insulin needs some caution yet, as they were sensitive to the removal of a particular included study.
Conclusion
Compared to other edible oils, rapeseed oil significantly reduced LDL-C, ApoB, ApoB/ApoA1 and insulin levels, and increased blood glucose levels in people with overweight and obesity. However, compared to other oils, rapeseed oil does not provide additional beneficial effects on weight and body composition in individuals with overweight and obesity. In the future, randomized controlled trials with large sample sizes, long intervention durations and high quality are needed to further validate the beneficial effects of rapeseed oil on weight and body composition in patients with obesity and overweight.
ARYAN'S BLOG 