first study on pubmed is the most relevant.
Effect of corosolic acid on gluconeogenesis in rat liver.Yamada K, Hosokawa M, Fujimoto S, Fujiwara H, Fujita Y, Harada N, Yamada C, Fukushima M, Ueda N, Kaneko T, Matsuyama F, Yamada Y, Seino Y, Inagaki N.
Department of Diabetes and Clinical Nutrition, Graduate School of Medicine, Kyoto University, 54 Shogoin, Kawahara-cho, Sakyo-ku, Kyoto 606-8507, Japan; Use Techno Corporation Co. Ltd., Kyoto, Japan.
Corosolic acid (CRA), an active component of Banaba leaves (Lagerstroemia speciosa L.), decreases blood glucose in diabetic animals and humans. In this study, we investigated the mechanism of action of CRA on gluconeogenesis in rat liver. CRA (20-100muM) dose-dependently decreased gluconeogenesis in perfused liver and in isolated hepatocytes. Fructose-2,6-bisphosphate (F-2,6-BP), a gluconeogenic intermediate, plays a critical role in hepatic glucose output by regulating gluconeogenesis and glycolysis in the liver. CRA increased the production of F-2,6-BP along with a decrease in intracellular levels of cAMP both in the presence and in the absence of forskolin in isolated hepatocytes. While a cAMP-dependent protein kinase (PKA) inhibitor inhibited hepatic gluconeogenesis, the drug did not intensify the inhibitory effect of CRA on hepatic gluconeogenesis in isolated hepatocytes.
These results indicate that CRA inhibits gluconeogenesis by increasing the production of F-2,6-BP by lowering the cAMP level and inhibiting PKA activity in isolated hepatocytes. Furthermore, CRA increased glucokinase activity in isolated hepatocytes without affecting glucose-6-phosphatase activity, suggesting the promotion of glycolysis. These effects on hepatic glucose metabolism may underlie the various anti-diabetic actions of CRA.
PMID: 18177973 [PubMed - as supplied by publisher]
I couldn't find any human studies. This goes back to the question of metabolism in rats vs. humans. Which has been debated to increasing degree around here.
this is form oxford and is more relevant:
The leaves of Lagerstroemia speciosa (Lythraceae), a Southeast Asian tree more commonly known as banaba, have been traditionally consumed in various forms by Philippinos for treatment of diabetes and kidney related diseases. In the 1990s, the popularity of this herbal medicine began to attract the attention of scientists worldwide. Since then, researchers have conducted numerous in vitro and in vivo studies that consistently confirmed the antidiabetic activity of banaba. Scientists have identified different components of banaba to be responsible for its activity. Using tumor cells as a cell model, corosolic acid was isolated from the methanol extract of banaba and shown to be an active compound. More recently, a different cell model and the focus on the water soluble fraction of the extract led to the discovery of other compounds. The ellagitannin Lagerstroemin was identified as an effective component of the banaba extract responsible for the activity. In a different approach, using 3T3-L1 adipocytes as a cell model and a glucose uptake assay as the functional screening method, Chen et al. showed that the banaba water extract exhibited an insulin-like glucose transport inducing activity. Coupling HPLC fractionation with a glucose uptake assay, gallotannins were identified in the banaba extract as components responsible for the activity, not corosolic acid. Penta-O-galloyl-glucopyranose (PGG) was identified as the most potent gallotannin. A comparison of published data with results obtained for PGG indicates that PGG has a significantly higher glucose transport stimulatory activity than Lagerstroemin. Chen et al. have also shown that PGG exhibits anti-adipogenic properties in addition to stimulating the glucose uptake in adipocytes. The combination of glucose uptake and anti-adipogenesis activity is not found in the current insulin mimetic drugs and may indicate a great therapeutic potential of PGG.
the full study can be found here:
http://ecam.oxfordjournals.org/cgi/content/full/nem013v1they give a history of discoveries, but all seem to be based on mice/rat studies. Seems like something to experiment with especially if your doing CKD. Getting back into keto could be very easy if it does the same in humans as in rats.