To investigate the potential of acetic acid supplementation as a means of lowering the glycaemic index (GI) of a bread meal, and to evaluate the possible dose–response effect on postprandial glycaemia, insulinaemia and satiety.
Subjects and setting:
In all, 12 healthy volunteers participated and the tests were performed at Applied Nutrition and Food Chemistry, Lund University, Sweden.
Three levels of vinegar (18, 23 and 28 mmol acetic acid) were served with a portion of white wheat bread containing 50 g available carbohydrates as breakfast in randomized order after an overnight fast. Bread served without vinegar was used as a reference meal. Blood samples were taken during 120 min for analysis of glucose and insulin. Satiety was measured with a subjective rating scale.
A significant dose–response relation was seen at 30 min for blood glucose and serum insulin responses; the higher the acetic acid level, the lower the metabolic responses. Furthermore, the rating of satiety was directly related to the acetic acid level. Compared with the reference meal, the highest level of vinegar significantly lowered the blood glucose response at 30 and 45 min, the insulin response at 15 and 30 min as well as increased the satiety score at 30, 90 and 120 min postprandially. The low and intermediate levels of vinegar also lowered the 30 min glucose and the 15 min insulin responses significantly compared with the reference meal. When GI and II (insulinaemic indices) were calculated using the 90 min incremental area, a significant lowering was found for the highest amount of acetic acid, although the corresponding values calculated at 120 min did not differ from the reference meal.
Supplementation of a meal based on white wheat bread with vinegar reduced postprandial responses of blood glucose and insulin, and increased the subjective rating of satiety. There was an inverse dose–response relation between the level of acetic acid and glucose and insulin responses and a linear dose–response relation between acetic acid and satiety rating. The results indicate an interesting potential of fermented and pickled products containing acetic acid.
Dr P Håkansson's foundation and Direktör Albert Påhlsson's foundation for research and charity.
Today, we see a rapid increase in obesity and diseases related to the insulin resistance syndrome (IRS). The health costs will rise profoundly, and it is a great challenge for the future to find means to counteract this development. The quality of the diet has been shown to play an important role in the combat of metabolic disorders and one quality parameter of dietary carbohydrates relates to the glycaemic index (GI), which is used to classify the glycaemic responses to carbohydrate rich foods (Jenkins et al, 1981). At this time point, there is a substantial amount of evidence that a diet characterized by a low GI has benefits in both prevention and treatment of several diseases linked to the IRS, such as cardiovascular disease (CVD) (Liu et al, 2000) and type II diabetes (Salmerón et al, 1997a, 1997b). New cross-sectional data also indicate that such a diet is associated with a lower prevalence of insulin resistance and metabolic syndrome (McKeown et al, 2004). Furthermore, evidences are at hand suggesting a beneficial role of low-GI foods adjunct to appetite regulation. Consequently, studies have demonstrated either increased satiety, delayed return of hunger or decreased ad libitum food intake after low-GI compared with high-GI foods (Ludwig, 2000). However, the role of dietary GI in relation to the increasing prevalence of overweight and obesity remains unclear. A shortcoming regarding the implementation of the GI concept is the lack of low-GI products on the food market.
A range of factors is known to affect the GI of carbohydrate-rich foods connected either to the characteristics of the raw material (ratio of amylose/amylopectin (Granfeldt et al, 1995), soluble fibre content (Braaten et al, 1991), inclusion of intact cereal kernels (Liljeberg & Björck, 1994), etc.) or the process (eg manufacturing process of pasta (Granfeldt et al, 1991), pumpernickel baking (Åkerberg et al, 1998) and fermentation (Liljeberg et al, 1995)). Previous results have shown that the presence of lactic acid lowers the GI of bread (Liljeberg et al, 1995) as well as improves the glucose tolerance at a subsequent high-GI meal (second-meal effect) (Östman et al, 2002a). The mechanism for the glucose-lowering action of lactic acid has been suggested to be due to a lowered rate of starch hydrolysis in the upper small intestine (Östman et al, 2002b). Other organic acids that have been investigated in relation to GI include acetic and propionic acids. Both these acids have shown to lower the glucose response to bread meals (Liljeberg & Björck, 1998; Darwiche et al, 2001), but in contrast to lactic acid, the mechanism of action appears to be a lowered rate of gastric emptying (Liljeberg & Björck, 1998; Darwiche et al, 2001). Propionic acid is present in certain types of cheese and is also used in low amount as a preservative, for example, in bread products. Acetic acid is found in a range of food products such as vinegar, dressings and pickled products. It is also formed upon fermentation processes, such as sourdough fermentation, and can be considered a constituent of a normal diet. The public interest for vinegar has been increasing in the past few years, and new products are currently being introduced in the market. In the present study, the possible dose–response relationship between the amount of acetic acid added to a bread meal in the form of vinegar, and the glycaemic and insulinaemic responses were evaluated in healthy subjects. In addition, the satiating effect of the different amounts of vinegar as supplements to a white bread meal was evaluated. The lower level of vinegar used in the present study is the same as the one used in a former study by Liljeberg and Björck (1998).
Subjects and methods
Subjects and meals
A total of 12 healthy, nonsmoking volunteers, 10 women and two men, aged 22.9 (s.e.m.=0.5) y, with normal body mass indices (21.4±0.7 kg/m2) and without drug therapy participated in the study. Besides a reference portion of white wheat bread, three test meals containing an identical portion of white wheat bread were served with 18, 23 or 28 g white vinegar (6% acetic acid, Druvan, Eslöv, Sweden), which is equivalent to 18, 23 and 28 mmol acetic acid in the respective portions. The highest amount of acetic acid corresponds with approximately 30 ml vinegar of this type. The white wheat bread was baked according to Liljeberg and Björck (1994) and was identical to the reference bread normally used in the GI determinations in our laboratory. The bread was soaked in the portion of vinegar before ingestion, and the intake of vinegar was thus distributed over the meal. All meals contained 50 g available starch. The subjects were served the meals in random order at four separate occasions. The tests were performed approximately 1 week apart and commenced at the same time in the morning. All meals were consumed steadily and finished within 12–14 min. Water (150 ml) and 150 ml tea or coffee was served with each meal. The test subjects were allowed to choose between these drinks at the first occasion and then the same drink was retained through all the test meals.
Sampling and analysis
The subjects arrived at the laboratory in the morning after an overnight fast. A fasting blood sample was taken and the subjects feeling of hunger/satiety was rated on a subjective rating scale before the meal was served. The rating scale was bipolar and graded from −10, to represent extreme hunger, to +10, to explain extreme satiety (Liljeberg et al, 1995). After the breakfast, blood samples were taken at 15, 30, 45, 60, 90 and 120 min for analysis of glucose and at 15, 30, 45, 90 and 120 min for analysis of insulin. The feeling of hunger/satiety was rated at 15, 30, 45, 70, 90 and 120 min postprandially. Blood glucose concentrations were determined with a glucose oxidase peroxidase reagent and serum insulin concentrations were determined with an enzyme immunoassay kit (Mercodia AB, Uppsala, Sweden).
The Ethics Committee of the Faculty of Medicine at Lund University approved the study.
The areas under the curves (AUCs) were determined for the blood glucose, serum insulin and satiety score (GraphPad Prism ver. 3.0; GraphPad Software, San Diego, USA). GI and II (insulinaemic indices) were determined by calculating the difference between the AUCs for the test meal and the reference meal in percent, with each subject being their own reference. All areas below the baseline were excluded from the calculations. Values are presented as mean±s.e.m. Subjects that rated their postprandial hunger lower than the hunger at fasting, hence indicating no satiety after the meal, were excluded from the calculations. All statistical calculations were performed in MINITAB Statistical Software (release 13 for Windows; Minitab Inc., State College, PA, USA). Significances were evaluated with the general linear model (analysis of variance) followed by Tukeýs multiple comparisons test. Values of P<0.05 were considered significant. The dose–response relation between the level of acetic acid and glucose, insulin or satiety AUC was evaluated with linear regression in GraphPad Prism (ver. 3.0).
When vinegar was served with the white wheat bread, the glycaemic responses were significantly lower at 30 min postprandially (P<0.05), compared with the reference meal without vinegar (Figure 1). No differences were noted in between the test meals with acetic acid at this time point. At 45 min, the highest amount of vinegar still lowered the glucose level (P=0.0540), compared with the reference. The GI of the test meal with the highest amount of vinegar (GI=77) was significantly lower than that of the reference meal, using the 90 min areas for calculation (Table 1). When GI was calculated using the AUC at 120 min no significant differences were found between any of the meals. A negative linear relation was found between 30 min blood glucose levels and vinegar content of the test meal (r=−0.47, P=0.001).
The insulin responses at 15 min were significantly lower for the high and intermediate levels of vinegar compared with the reference meal (Figure 2). At 30 min, only the highest amount of vinegar lowered the insulin level significantly compared with the reference. The II of the meal with highest amount of vinegar (II=78) was significantly lower than the reference, when using the 90 min AUCs for calculation (Table 2). Calculating with the AUCs at 120 min, no significant differences in II were found between any of the meals. When expressing the 30 min insulin responses as a function of the vinegar content of the test meal, a negative linear relation was found (r=−0.44, P=0.002).
The reference meal resulted in the lowest rating of satiety (Figure 3). The satiety scores after the highest level of vinegar were rated significantly higher than the reference at 30, 90 and 120 min after the meal. The score for the low and intermediate levels of vinegar did not differ significantly at any of the time points from either the reference or the meal with a high amount of vinegar. The satiety AUC was significantly larger for the meal with a high amount of vinegar compared with that of the reference meal (Table 3). A significant linear relation was found between the satiety AUC and the acetic acid content of the test meals (r=0.41, P=0.004). The satiety rating of the reference meal without acetic acid had returned to the rating at fasting after 90 min, whereas the meals with acetic acid maintained positive ratings over the entire experimental period of 120 min (Figure 3).
The most important finding in the present study was that addition of acetic acid/vinegar improved not only the postprandial blood glucose and insulin profiles, but also the satiating effect of a meal based on white wheat bread. With respect to satiety rating, the intake of acetic acid/vinegar appeared not only to increase, but also to prolong satiety. Furthermore, a significant dose–response relation was found at 30 min postprandially for blood glucose and serum insulin responses, indicating that the higher the acetic acid level, the lower the metabolic response.
In previous work by Liljeberg and Björck (1998) with meals based on white wheat bread, a 35% lowering of GI and II was seen with 18 mmol acetic acid (20 g vinegar) when served as a vinaigrette sauce. In the present study, the same amount of acetic acid (18 mmol) lowered mean GI by 11% but without reaching statistical significance, and with no effect on II. In the study by Liljeberg and Björck (1998), the vinegar was supplied as a vinaigrette sauce including water and olive oil (8 g). Results from studies investigating the effect of fats with various degrees of saturation on the blood glucose response of a meal are conflicting (Gatti et al, 1992; MacIntosh et al, 2003). Consequently, it remains to be elucidated if the combination of acetic acid and olive oil caused the more potent lowering of the glycaemic response in the work by Liljeberg and Björck (1998). However, in a recent study, the addition of 15 g sunflower oil (62% PUFA, 26% MUFA and 12% SFA) had no effect on the glycaemic excursions following a potato meal in healthy subjects (Johnsson, Andersson, Granfeldt and Björck, unpublished data). Also, in work by MacIntosh et al (2003) no differences in GI or II were found for a meal with mashed potato and either butter (SFA), sunola oil (MUFA) or sunflower oil (PUFA). Furthermore, Owen and Wolever (2003) concluded that a variation of fat intake (nonhydrogenated-fat margarine: 41% PUFA, 41% MUFA and 14% SFA) across the normal range did not significantly affect the glycaemic response to a white bread meal.
The glucose-lowering effect of acetic acid/vinegar has also been observed by other investigators. Johnston et al (2004) reported that vinegar could significantly improve insulin sensitivity in insulin resistant subjects. In a recent study by Sugiyama et al (2003), the GI of a white rice meal was lowered with 25–35% by adding either vinegar (11 g/portion) or pickled cucumber (15 g vinegar/portion). Furthermore, Brighenti et al (1995) found a 30% lowering of AUC for glucose of 17 mmol acetic acid when served as a vinaigrette together with sliced lettuce and white bread. In the latter study, the suggested glucose-lowering effect was an inhibition of digestive amylases caused by the acid. However, in vitro measurements of the rate of starch hydrolysis using a simulated gastrointestinal model did not reveal any amylase inhibition in the case of acetic acid containing bread (Liljeberg et al, 1996). Moreover, as judged from studies in healthy volunteers using paracetamol as a marker for the gastric emptying rate, the presence of acetic acid appeared to significantly reduce the gastric emptying rate, suggesting an effect different from obstructed amylolysis (Liljeberg & Björck, 1998). In a study by Ebihara and Nakajima (1988) the administration of acetic acid improved the glycaemia in rats and lowered the insulin response in humans significantly. No lowering effect on the glucose response was seen in humans, but a postponed glucose peak and prolonged time to reach the fasting level postprandially were registered with the acetic acid containing meal. Also, in the rat study, the glucose response showed a slower decrease in the late postprandial phase in the presence of the acid, which is in line with a mechanism related to a lowered gastric emptying rate (Ebihara & Nakajima, 1988).
In studies from the 1960s and 1970s, the influence of organic acids on gastric emptying was evaluated in both humans (Hunt & Knox, 1969, 1972) and animals (Blum et al, 1976). In one of these early reports, Hunt and Knox (1972) stated that the higher the molecular weights of weak acids, the lower the potential of slowing gastric emptying. This finding suggests that acetic acid (Mw=60 g/mol) would be more effective than lactic acid (Mw=90 g/mol) in lowering the gastric emptying rate. This is in accordance with results from Liljeberg and Björck (1996, 1998) showing that the presence of acetic acid but not lactic acid lowered the postprandial appearance of paracetamol in the blood, indicative of a delay in gastric emptying rate in the case of acetic acid. In addition to organic acids, the rate of gastric emptying can also be influenced by volume, caloric content, viscosity, density and particle size of the gastric content (Horowitz et al, 1994). Some peptides that are released in the small intestine, such as cholecystokinin (CCK), amylin and glucagon-like peptide-1 (GLP-1), have also been shown to be involved in the regulation of the gastric emptying rate (Hellström & Näslund, 2001; Feinle et al, 2002).
A difference between lactic- and acetic acid supplemented bread meals, beyond their immediate glucose-lowering effect, is that improved glucose tolerance at a subsequent standardized meal (second meal effect) has only been observed in meals with lactic acid (Östman et al, 2002a). When acetic acid was served as a supplement to a white bread breakfast meal, no effect on the glucose tolerance was observed at the subsequent standardized lunch meal (Liljeberg et al, 1999). In contrast, the presence of lactic acid in a barley bread-based breakfast meal reduced glycaemia at a subsequent standardized lunch meal by 25% (Östman et al, 2002a). The suggested mechanism for a second meal effect of a low-GI food is that the prolonged digestive phase causes a suppression of the plasma levels of free fatty acids (Wolever et al, 1995). This may lead to improved insulin sensitivity at the time of the next meal. The difference in the second meal effect between low-GI foods may be related to the fact that different gastrointestinal mechanisms may be responsible for the slow release properties. Since a lowered gastric emptying, and not a lowered starch digestion, causes the glucose-lowering effect of acetic acid, it can be hypothesized that the criteria for a second meal effect in the perspective from breakfast to lunch depend mainly on a mechanism connected to a prolonged digestive phase. In animal studies, Fushimi et al (2001) showed that acetic acid could activate gluconeogensis and induce glycogenesis in the liver after a fasting state. Furthermore, when Caco-2 cells (human colonic carcinoma cells) were cultivated with acetic acid for 15 days, Ogawa et al (2000) showed that acetic acid suppresses the increase in disaccharidase activity (sucrase, maltase, trehalase and lactase). These two latter studies add pieces of evidence to the knowledge on how acetic acid affects glucose absorption and metabolism, but more research is needed to understand the importance of each physiological phenomenon.
There is currently an ongoing discussion concerning the need to standardize the GI methodology. One potentially interesting issue for ranking of carbohydrate characteristics relates to the time interval used for the calculation of GI. In the present study, there are important differences between glucose and insulin responses to the vinegar containing meals and the reference meal in the early postprandial phase (15–30 min). Despite these lowering effects of vinegar, a difference in GI and II is only significant for the meal with the highest amount of vinegar at 90 min and not at 120 min. This suggests that the ranking of products may be slightly different depending on the time period used. In fact, products with an extended digestive phase and with low but sustained increments in blood glucose may produce a substantial late glycaemic area, thus evening out differences in incremental areas compared with the rapidly digested and absorbed starch in the wheat bread reference. This has previously been shown to be the case with pasta (Granfeldt et al, 1991).
In the present study, the meal with the highest amount of vinegar also increased the late postprandial satiety more than two-fold, compared with the reference meal. None of the former articles on meal supplementation with vinegar have evaluated satiety, so the current finding adds new knowledge both to the specific effect of vinegar on satiety and to the more general relation between GI and satiety, suggesting that low-GI foods can produce not only a higher satiety, with potential effects on food intake, but also prolong the duration of satiety, which may influence the voluntary intake at a subsequent meal.
Significant dose–response relations between the amount of added vinegar and the glucose, insulin or satiety responses were found based on the 30 min values in the present study. The highest amount of acetic acid used in the present study corresponds to approximately 30 ml of vinegar. This is a level comparable with that of acetic acid in a portion of pumpernickel bread (GI=68, II=63) previously tested in our laboratory (Liljeberg & Björck, 1994). However, a similar amount of acetic acid served as a salad dressing, as pickled vegetables may be difficult to ingest. The selection of pickled and fermented products or meal additives, and the use of vinegar-based drinks, which are currently introduced in the market, may provide means to reach efficient levels of acetic acid. Addition of vinegar to carbohydrate-rich meals of high-GI character, or the use of, for example, homofermentative, acetic acid producing starter cultures offers a potential to lower the GI and increase the postmeal satiety. The possible long-term health benefits of including pickled products or fermented products in the diet need to be examined.