The two fruit drinks in series 1, BFOMD and RFOMD, gave a postprandial blood glucose response similar to that after an equivalent amount of carbohydrate from white bread. The GI was thus 97 and 89 for BFOMD and RFOMD, respectively. However, the high blood glucose responses were not accompanied by corresponding high insulin responses. Consequently, a tendency to a lower insulin response was present after both fruit drinks compared to that after white bread, even though only the area under the insulin curve after the BFOMD (0-120 min), was significantly smaller than after white reference bread. The II was determined to 65 (P < 0.05) and 79 (P = 0.0673) for BFOMD and RFOMD, respectively.
When calculating GI's, according to the content of digestible carbohydrates and their GI values (table ), the BFOMD received a GI of 95 and the RFOMD a GI of 97. Accordingly, the determined GI values in series 1 (97 and 89 for BFOMD and RFOMD respectively) are in good agreement with those calculated. Previously, also Gannon et al
] found that glucose response to fruits (oranges and apples) can be predicted from the constituent carbohydrates present, whereas the insulin response cannot. However, in contrast to our finding with low insulin responses Gannon et al found higher observed insulin responses than predicted from glycaemia in the case of orange- and apple juice.
Available carbohydrate composition, and calculated GI (Series 1 and 2)
In series 2, the GI and II for a BFOMD, matching that in series 1, were determined to 92 and 64, respectively compared with GI = 97 and II = 65 in series 1. Thus, the favourable effect of a fermented oat meal drink with bilberry on insulin demand in series 1 could be repeated. The FOMD gave a high GI (GI = 95) whereas that of the BBFOMD was lower (GI = 79). Insulin indices (II) for the FOMD and the BBFOMD were; II = 76 and 49, respectively. The II for the BBFOMD was remarkably low, consequently when using white bread as a reference, beverages like soft drinks (II = 97-118) [19
], and other fruit/berry based drinks like orange-(II = 78) [19
] or apple juice (II = 54, (estimated from insulin areas) [3
] BBFOMD has a low II. Apple juice and BBFOMD resembling, that after e.g. pasta products (II = 35-53) [22
The content of carbohydrates in the different products in the present study was similar, the amount of sugars being high (approximately 90%) and that of starch low. When calculating GI values according to the content of digestible carbohydrates and their GI values (table ), the three test drinks in series 2, FOMD, BFOMD and BBFOMD received; 95, 95 and 90 respectively, to be compared with the determined GI values; 95, 94 and 79. The observed GI value for the BBFOMD was thus lower than expected; but for the other drinks the calculated and determined GI values were in good agreement. The decreased acute glycaemic response with a larger amount of bilberries (145.5 g) is noteworthy, and a hypothesis may be that the bilberries cause an increased uptake of glucose into the peripheral cells. Thus, Martineau et al
] found insulin-like properties of ethanol extracts from Canadian low bush blueberries (Vaccinium angustifolium)
, another member of the Vaccinium
genus. The insulin-like properties were evident from enhancement of glucose uptake in differentiated muscle cells and adipocytes using an in vitro assay. Bilberry extract was also shown to reduce blood glucose level and enhance insulin sensitivity in type 2 diabetic mice via activation of AMPK (AMP-activated protein kinase), an enzyme central in the regulation of fuel preference, in adipose tissue, muscle- and liver cells [16
]. These results support our finding of a decreased acute glycaemic response with bilberry and show that bilberry (Vaccinium myritillus
) may contain active molecules with potential anti-diabetic properties. Such an effect, if present, is also coherent with the lowered insulin demand with bilberry drinks seen in both series 1 and 2. The existence of a dose-response relation between intake of bilberries and the corresponding glucose/insulin responses is currently in progress.
In this study we show an inconsistency between glycaemic and insulinaemic responses, especially in the case of the products containing bilberries. Consequently a high glucose response was accompanied by a comparatively low insulin response. The insulin demand, if expressed as a relationship between insulinaemia and glycaemia (II/GI), was low, 0.62-0.68. To our knowledge, this apparent discrepancy with a low insulin response in parallel to a high glucose response has only been reported previously for fermented whole-grain oat [23
], certain rye products [24
] and for cinnamon added to a rise pudding [25
]. Earlier studies generally have shown a good correlation between glucose and insulin responses. Thus, studies with cereals [26
], and certain fruits like mango, melon, pineapple, kiwi, apple and black grapes [19
] indicate good agreement between GI and II. In contrast studies with oranges and apples [28
] as well as juice from these fruits [3
] were reported to display unexpectedly high insulin responses. Similarly, a discrepancy between GI and II, with unexpectedly high insulin responses has been shown, for milk and milk products [30
]. Consequently despite extremely low GI (GI = 15-30) for regular and fermented milk, the II values were high (II = 90-98) [31
], probably due to an insulinotrophic effect of whey protein [33
]. One cause for the beneficial metabolic effects of a low glycaemic diet is probably a lower insulin response [34
], and increased insulin sensitivity [36
]. The present findings of low insulin demand following bilberry drinks might thus indicate advantageous metabolic properties.
The fermented oat meal base of the drinks, with sugars added to mimic the BFOMD was included as a reference drink (FOMD). A comparison between the FOMD and the BFOMD, gave no significant differences in glucose or insulin responses. Also, the insulin response to the FOMD (II = 76) was significantly lower than for white bread (II = 100). In a previous study with oats (oat porridge and oat flakes) no differences in glucose- or insulin responses were seen compared with a white bread [37
]. This indicates that the fermentation process per se may decrease insulin response to oats. It is also supported by a study with fermented whole grain oat showing a lower insulin response than would be expected from the glucose values [23
]. However, the magnitude of insulin decrease in the post prandial phase was more pronounced when more bilberries were included in the meal. Of interest in this respect are results from an in vitro study evaluating the effect of fermented blueberry juice (intrinsic micro flora of blueberries) on glucose uptake and transport into muscle cells and adipocytes. Treatment of cells with fermented juice potentiated glucose uptake by 48% in C2C12 (mouse myoblast cell line) myotubes, and by 142% in 3T3-L1 adipocytes, whereas non fermented juice had no effect on glucose transport [38
]. Treatment of cells with fermented blueberry juice was shown to activate AMPK. The authors thus suggest an insulin-independent pathway to be the mechanism for an increased glucose uptake [38
]. A follow-up study in obese and diabetic mice showed that fermented blueberry juice decreased hyperglycaemia, in part due to increased adiponectin levels. However, no positive effects were seen on insulin levels [17
]. In the presently reported study, we saw a decreased early glucose response after BBFOMD. However, insulin responses were significantly lower, or close to being significantly lower for the drink with rose hip (RFOMD) (P = 0.0673), and for all fermented test drinks whether containing bilberries or not. Whether the low insulin demand (II/GI) shown in the present study could be an effect of fermentation of bilberries is currently under investigation.
All tested drinks had a low pH, or about 3 for the bilberry containing drinks, and about 4 for the fermented oat meal drink (FOMD). A low pH may lower post prandial glycaemia and hormonal responses due to e.g. a lowering of the rate of gastric emptying [39
]. However, such a mechanism should preferably affect both blood glucose and insulinaemia to a similar extent.
Berries like bilberries and blueberries are known to be a rich source of bioactive molecules like phenolic and antocyanin contents [40
] and phenolic acids [41
]. Besides that they both are powerful antioxidants they may also exert effects on other important biological systems as glucose- and insulin response. Bilberries mixed with blackcurrants, cranberries and strawberries (150 g), other berries rich in antocyanins, have recently been shown to decrease the peak glucose increment of 35 g sucrose in healthy subjects [18
]. Also in type-2 diabetic mice, anthocyanins in bilberry have been suggested to reduce blood glucose levels and enhance insulin sensitivity [16
]. Water soluble polyphenols isolated from cinnamon has been shown to have strong insulin-enhancing activity on cultured fat cells in vitro, [42
]. Also, when tested in healthy subjects, cinnamon (3 g) added to a rice pudding (300 g) was shown to reduce post prandial serum insulin, but not glucose, levels compared to a rice pudding without cinnamon [25
]. It is impossible to draw any conclusions regarding the effects of the antocyanins and/or polyphenols present in the oat meal based fruit drinks (RFOMD, BFOMD, BBFOMD) in the present study. However, it cannot be excluded that such components might have contributed to the low insulin demand seen after the fermented oat meal drink added with bilberries.
To our knowledge no meal studies have been published showing impact of bilberries on glycaemic and insulinaemic responses in humans.