Canagliflozin

CANVAS Program, 2017

• The program combined two randomized placebo-controlled trials, CANVAS and CANVAS-R, in 10,142 adults with type 2 diabetes and high cardiovascular risk (mean age 63.3 years, 35.8 percent women, 65.6 percent with a history of cardiovascular disease), who were followed for a mean of 188.2 weeks (PMID: 28605608).
• The program met its primary outcome: cardiovascular death, nonfatal heart attack or nonfatal stroke occurred at 26.9 against 31.5 per 1,000 patient-years (hazard ratio 0.86, 95% CI 0.75 to 0.97; P=0.02 for superiority) (PMID: 28605608).
• Under the prespecified testing order the kidney outcomes were not counted as statistically significant, although the results suggested less progression of albuminuria (hazard ratio 0.73) and fewer composite kidney events (hazard ratio 0.60) (PMID: 28605608).
• Amputation risk was higher with canagliflozin (hazard ratio 1.97), as described under side effects (PMID: 28605608).

CREDENCE, 2019

• CREDENCE was a double-blind randomized trial of canagliflozin 100 mg a day or placebo in 4,401 adults with type 2 diabetes and albuminuric chronic kidney disease (eGFR 30 to below 90 mL/min/1.73 m2, urine albumin-to-creatinine ratio above 300 to 5,000 mg/g) who were taking renin-angiotensin system blockers. The trial was stopped early after a planned interim analysis on the recommendation of its data and safety monitoring committee, at a median follow-up of 2.62 years (PMID: 30990260).
• The trial met its primary outcome: end-stage kidney disease, a doubling of serum creatinine, or death from renal or cardiovascular causes occurred at 43.2 against 61.2 per 1,000 patient-years (hazard ratio 0.70, 95% CI 0.59 to 0.82) (PMID: 30990260).
• Among secondary outcomes, the kidney-specific composite (hazard ratio 0.66), end-stage kidney disease (0.68), cardiovascular death, heart attack or stroke (0.80) and hospitalization for heart failure (0.61) were all lower with canagliflozin, with no significant differences in amputation or fracture (PMID: 30990260).

Glucose lowering

A 2017 review found that canagliflozin, alone or added to other diabetes drugs, improved glycemic control in adults with type 2 diabetes (among them older adults and people at high cardiovascular risk), with reductions in body weight and blood pressure and a low risk of hypoglycemia (PMID: 28836175). On 18 December 2024 the FDA extended the glycemic indication to children aged 10 years and older, based on a placebo-controlled pediatric trial (FDA approval letter, NDA 204042 supplement 43).

A registered study in healthy adults

A 2-week, non-randomized, open-label study of canagliflozin in 30 healthy adults, registered as a test of canagliflozin as a gerotherapeutic with continuous glucose monitoring as its primary outcome, was completed in 2024 and posts no results (ClinicalTrials.gov NCT06301529).

Interventions Testing Program lifespan studies (animal data)

• In the 2016 cohort, published in 2020, canagliflozin at 180 ppm in the diet from 7 months of age, given to about 50 mice of each sex at each site, extended male median lifespan by 14 percent and male 90th percentile age by 9 percent, with parallel effects at each of the 3 sites; female lifespan did not change, and the inferred causes of death did not shift (PMID: 32990681). Treated females weighed 17 to 19 percent less than controls at each age measured after the start of treatment, treated males weighed 5 to 8 percent less at 12 and 18 months, and HbA1c did not change in either sex (PMID: 32990681).
• In the 2020 cohort, published in 2024, canagliflozin started at 16 months of age, in a group run at two of the three sites (Michigan and UT Health San Antonio), increased male median lifespan by 14 percent (p=0.004) and decreased female median lifespan by 6 percent (p=0.03); a group started at 6 months, run only at The Jackson Laboratory after a technical error, again lengthened male median lifespan, by 9 percent, without changing female lifespan, and the authors raised the possibility of female-specific late-life harm (PMID: 38753230). A 2025 analysis of end-of-life pathology in these mice found no clear pattern that explained the sex-specific effects (PMID: 40601216).
• A 2025 review of the 54 agents tested by the ITP found that most compounds that extended lifespan worked mainly or only in male mice (PMID: 40717358). BodyHackGuide's rapamycin and metformin guides cover two other drugs the program has tested.

Other mouse studies (animal data)

• In 30-month-old mice, canagliflozin improved brain insulin sensitivity in males and reduced hippocampal glial activation in males only (PMID: 35707855). A 2025 study found better hippocampus-dependent learning and memory in aged male UM-HET3 mice but not females, and in 5xFAD Alzheimer's model mice, canagliflozin reduced amyloid plaque, microgliosis and memory deficits in males only (PMID: 41047765).
• UM-HET3 mice fed canagliflozin from 7 to 22 months of age had a thinner femur cortex in both sexes and lower cortical bone mineral density in males (PMID: 37166526).

Lower limb amputation

In the CANVAS Program, amputations occurred at 6.3 against 3.4 per 1,000 participant-years with canagliflozin and placebo (hazard ratio 1.97, 95% CI 1.41 to 2.75), mainly at the level of the toe or metatarsal (PMID: 28605608). A secondary analysis of the 187 people with amputations found similar risk at 100 and 300 mg and for ischemic and infective causes, with a previous amputation as the strongest risk factor (hazard ratio 21.31) (PMID: 30868176). The CREDENCE trial found no significant difference in amputation rates (PMID: 30990260). A pooled analysis of the two programs could not explain the difference between them and concluded that the CANVAS finding may have been due to chance, given null results in CREDENCE and in other large SGLT2 inhibitor trials (PMID: 32436638). The current label keeps lower limb amputation as a warning and cites 5.9 against 2.8 events per 1,000 patient-years in CANVAS and 7.5 against 4.2 in CANVAS-R (US prescribing information for Invokana).

Bone fracture

In the CANVAS Program, fractures occurred at 15.40 against 11.93 per 1,000 patient-years (hazard ratio 1.26, 95% CI 1.04 to 1.52), with a higher risk in CANVAS (hazard ratio 1.55) but not in CANVAS-R (hazard ratio 0.86); the authors could not explain the difference and, given the null CREDENCE result, judged the CANVAS association likely to be a chance finding (PMID: 31399845, PMID: 30990260). The US label keeps a bone fracture warning and notes fractures as early as 12 weeks after the start of treatment in CANVAS (US prescribing information for Invokana). In ITP mice given canagliflozin from 7 to 22 months of age, the femur cortex became thinner in both sexes and cortical bone mineral density fell in males (PMID: 37166526).

Ketoacidosis, kidney injury and genital infections

The label warns of ketoacidosis (including fatal postmarketing cases in type 2 diabetes), with blood glucose at presentation that may be below 250 mg/dL; of volume depletion, with postmarketing reports of acute kidney injury, some needing hospitalization and dialysis; of genitourinary infections including urosepsis, pyelonephritis and necrotizing fasciitis of the perineum (Fournier's gangrene); of hypoglycemia when combined with insulin or insulin secretagogues; and of hypersensitivity reactions (US prescribing information for Invokana). An FDA review of spontaneous reports identified 55 cases of Fournier's gangrene in people taking SGLT2 inhibitors, the drug class that includes canagliflozin, between 1 March 2013 and 31 January 2019; all needed surgical debridement and 3 died (PMID: 31060053).

Female mice in late life (animal data)

When the ITP started canagliflozin at 16 months of age, female lifespan fell by 6 percent, and end-of-life pathology showed no clear pattern that explained the sex-specific effects (PMID: 38753230, PMID: 40601216).