Cannabis Extract Fixes Glucose Impairment That Pure THC Cannot

A preclinical study from UC Riverside published May 11, 2026 in The Journal of Physiology provides new mechanistic evidence for why whole-plant cannabis produces metabolic benefits that isolated THC does not — and why that distinction matters for the long-standing epidemiological puzzle around cannabis and type 2 diabetes.

The headline result: both pure delta-9-THC and a whole-plant cannabis extract reduced body weight and fat mass in diet-induced obese mice. But only the extract normalized glucose clearance, reversing the impaired insulin response that typifies metabolic dysfunction. THC reduced weight, yet left glucose regulation broken.

Two Treatments, Two Metabolic Outcomes

The team, led by UCR Center for Cannabinoid Research director Nicholas DiPatrizio, fed male mice a high-fat, high-sugar Western-style diet for 60 days to induce obesity and metabolic dysfunction. At day 30, mice began receiving either pure THC at 5 mg/kg daily or a whole-plant cannabis extract matched for THC content — both groups received the same THC dose. The cannabis flower used in the study was donated by the National Institute on Drug Abuse and characterized by mass spectrometry: alongside 5 mg/kg of THC, the extract contained CBN, THCA, CBG, THCV, and trace CBD.

After 30 days of treatment, glucose tolerance testing separated the groups sharply. Mice on THC lost weight and shed visceral fat — but their glucose clearance remained statistically indistinguishable from untreated obese controls. The whole-plant extract group normalized glucose clearance to levels approaching lean mice on a standard diet. Both treatments reduced circulating insulin and fasting blood glucose, but only the extract restored insulin sensitivity and normalized the adipokine profiles disrupted by obesity.

Sample sizes are modest — five to nine animals per group for the metabolic endpoints — so these are signals, not settled numbers. The direction is consistent across multiple outcomes.

“This suggests that THC alone is not responsible for the metabolic benefits associated with cannabis use,” DiPatrizio said. “Other compounds in the plant appear to play a critical role.”

The Adipoinsular Axis: Fat Talking to the Pancreas

The mechanism the researchers point to is a signaling network called the adipoinsular axis. In healthy metabolism, fat cells secrete molecules — adipokines including leptin, adiponectin, and adipsin — that communicate with the pancreas and help regulate insulin secretion. In obesity, this signaling degrades: leptin levels rise and leptin resistance sets in, adiponectin falls, adipsin drops, and the crosstalk between adipose tissue and pancreatic beta cells breaks down.

The study found that whole-plant extract restored expression of these adipokines more completely than THC alone. Adipsin gene expression in visceral fat — which had fallen sharply in obese control mice — was partially restored by THC but more fully restored by the extract. Leptin expression was significantly reduced by the extract but not by THC alone. These gene-level changes tracked with plasma measurements and with the glucose tolerance results.

The endocannabinoid system itself also appeared to normalize. Elevated levels of 2-AG and anandamide — both of which rise in obesity and contribute to metabolic dysregulation — were reduced in extract-treated mice, consistent with chronic cannabinoid exposure driving downregulation of cannabinoid receptor expression in adipose tissue. That CB1 receptor-downregulation mechanism is the same process invoked to explain the “cannabis paradox” in human epidemiology: regular cannabis users are approximately half as likely to develop type 2 diabetes compared to non-users, despite acute cannabis exposure increasing appetite. The working hypothesis — supported here — is that chronic cannabinoid exposure resets endocannabinoid tone rather than permanently elevating it.

What the extract data adds is that resetting glucose homeostasis requires more than THC. The THCV, CBN, and CBD present in the extract, even at trace quantities relative to THC, appear to contribute something THC cannot provide alone. The researchers raise the possibility of synergistic or combinatorial interactions among multiple phytocannabinoids — the same whole-plant versus isolated compound question that recurs throughout cannabinoid research.

What This Doesn’t Show

The limitations are worth stating directly. This is a mouse study — male mice only, a meaningful gap given well-documented sex differences in endocannabinoid signaling and metabolic regulation. The model does not fully recapitulate the complexity of human obesity and metabolic syndrome, and direct measurement of whole-body energy expenditure was not performed, leaving the mechanism for sustained weight loss — whether altered nutrient partitioning, thermogenesis, or something else — incompletely resolved. Housing below the thermoneutral zone for mice may also have influenced the magnitude of observed metabolic responses.

The study used intraperitoneal injections at doses calibrated to approximate blood levels seen in humans after consuming a single cannabis cigarette, which is a reasonable translational approach but still a rodent-pharmacokinetics extrapolation.

DiPatrizio’s stated next step is to isolate individual non-THC compounds and test each separately, aiming to identify non-psychoactive candidates that could deliver metabolic benefits without THC’s intoxicating effects. That’s a tractable experimental question, and this study gives it a specific mechanistic framework: restore the adipoinsular axis.

The growing cannabinoid research literature on metabolic and inflammatory conditions has been building toward a compound-specific map of who does what in the cannabis plant. This paper adds a data point the field needed: THC appears to account for the weight loss, but the rest of the plant appears to account for glucose.