Understanding SGLT2 Inhibitors

Sodium‑glucose cotransporter 2 (SGLT2) inhibitors are a cornerstone of modern type 2 diabetes management. By selectively blocking the SGLT2 protein in the proximal renal tubule, these agents reduce glucose reabsorption and promote urinary glucose excretion. The resulting glucosuria lowers plasma glucose levels independently of insulin secretion, making SGLT2 inhibitors effective across a wide range of patients. Beyond glycemic control, the class confers substantial cardiovascular and renal benefits: reduced risks of major adverse cardiovascular events, hospitalization for heart failure, and progression of chronic kidney disease. The renoprotective effects have been robustly demonstrated in trials such as CREDENCE and DAPA‑CKD, leading to expanded indications for use even in non‑diabetic patients with CKD. The three most studied drugs—canagliflozin, dapagliflozin, and empagliflozin—are now first‑line agents in many clinical guidelines.

The Emergence of Bone Density Concerns

Despite their favorable profile, early signals from the CANVAS Program raised concern about SGLT2 inhibitors and fracture risk. In that trial, canagliflozin was associated with an increased incidence of bone fractures, particularly at non‑weight‑bearing sites such as the arm and hand. This unexpected finding prompted regulatory notifications, including an FDA Drug Safety Communication in 2015 warning about potential bone density loss with canagliflozin. Subsequent analyses of other SGLT2 inhibitors—including dapagliflozin and empagliflozin—did not show a statistically significant increase in fracture risk, but the question of whether the entire class or individual agents affect bone metabolism remained unresolved. As the use of SGLT2 inhibitors extends to older, frailer populations with higher baseline risks for osteoporosis, understanding the relationship between these drugs and bone health becomes clinically critical.

Research Findings on SGLT2 Inhibitors and Fracture Risk

Clinical Trial Evidence

The CANVAS Program, which randomized 10,142 patients to canagliflozin or placebo, reported a roughly 40% increase in fracture events (hazard ratio 1.4, 95% CI 1.0–1.9). The fractures occurred earlier and were predominantly at peripheral sites—not typical osteoporotic fractures of the hip or spine. In contrast, the EMPA‑REG OUTCOME trial with empagliflozin, the DECLARE‑TIMI 58 trial with dapagliflozin, and the VERTIS‑CV trial with ertugliflozin all showed no significant fracture signal. A 2022 meta‑analysis of 47 randomized controlled trials involving nearly 80,000 patients found that overall SGLT2 inhibitor use was not associated with a significantly higher fracture risk, but a subgroup analysis identified an increased risk only for canagliflozin. This heterogeneity suggests that the fracture risk may not be a class effect.

Observational Studies and Real‑World Data

Real‑world cohort studies have provided conflicting evidence. A large Danish registry study found no increase in fracture risk with SGLT2 inhibitors compared with GLP‑1 receptor agonists. Conversely, a Canadian population‑based study reported a modest increase in risk within the first 90 days of SGLT2 inhibitor initiation, driven mainly by canagliflozin users. Observational data are limited by confounding by indication—patients with more severe diabetes or vascular disease may have a higher baseline fracture risk. Nevertheless, the preponderance of evidence points to canagliflozin having a unique effect, while other agents appear neutral or even protective in some analyses.

Putative Mechanisms Linking SGLT2 Inhibitors to Bone Loss

Alterations in Serum Phosphate

One of the leading hypotheses involves changes in phosphate homeostasis. SGLT2 inhibitors increase urinary glucose excretion, and glucose in the tubule lumen promotes sodium and water excretion. This osmotic diuresis also increases phosphate reabsorption in the proximal tubule, leading to a modest rise in serum phosphate levels. Elevated phosphate can stimulate fibroblast growth factor 23 (FGF‑23) secretion, which in turn suppresses 1,25‑dihydroxyvitamin D production, reducing intestinal calcium absorption and potentially causing secondary hyperparathyroidism. Both a rise in PTH and low vitamin D can accelerate bone resorption. This mechanism has been demonstrated more consistently with canagliflozin than with other SGLT2 inhibitors, possibly due to its higher potency and longer half‑life.

Changes in Bone Remodeling Markers

Short‑term studies have measured markers of bone turnover during SGLT2 inhibitor therapy. Canagliflozin was associated with increases in C‑telopeptide (CTX), a marker of bone resorption, and decreases in procollagen type 1 N‑propeptide (P1NP), a marker of bone formation, leading to an imbalance favoring resorption. These changes correspond to a net bone loss of approximately 1–2% in bone mineral density (BMD) at the total hip and femoral neck over one year, as observed in a small substudy of the CANVAS trial. Dapagliflozin and empagliflozin have not shown such striking changes in bone markers or BMD in similarly designed studies.

Weight Loss and Hemoconcentration Effects

SGLT2 inhibitors promote modest weight loss (2–3 kg on average). While weight reduction is generally beneficial, it can slightly decrease areal BMD measured by DXA because of reduced soft tissue attenuation. This effect is unlikely to represent true bone loss, but it may contribute to signals seen in short‑term DXA studies. Additionally, volume contraction from diuresis causes hemoconcentration, which can increase the apparent density of bone measured by quantitative computed tomography. However, DXA is less affected by hydration status. Clinicians should interpret DXA changes cautiously in the context of simultaneous weight loss.

Potential Impact on Vitamin D and Calcium Metabolism

The observed rise in serum phosphate with SGLT2 inhibitors leads to FGF‑23 elevation, which reduces renal conversion of 25‑hydroxyvitamin D to 1,25‑dihydroxyvitamin D. Lower active vitamin D levels impair intestinal calcium absorption, potentially triggering a compensatory rise in PTH. Higher PTH can directly stimulate osteoclast activity. In the CANVAS BMD substudy, PTH levels increased significantly among canagliflozin users compared with placebo. Similar significant changes were not seen with other SGLT2 inhibitors, which may explain the class‑specific risk. Ongoing research is exploring whether calcium and vitamin D supplementation can mitigate these effects.

Clinical Implications for Diabetic Patients

Risk Stratification

Given the evidence, prescribers should assess fracture risk when initiating an SGLT2 inhibitor. Patients with a history of osteoporotic fracture, osteoporosis (T‑score ≤ −2.5), or high FRAX score are candidates for an agent with lower perceived bone risk—namely empagliflozin or dapagliflozin—rather than canagliflozin. Those with severe chronic kidney disease (stage 4 or 5) or hyperparathyroidism may be at heightened risk due to altered bone and mineral metabolism. Baseline BMD testing (DXA) should be considered in patients with multiple risk factors, especially if canagliflozin is chosen.

Monitoring During Therapy

For patients on canagliflozin, periodic assessment of bone markers such as serum PTH, calcium, phosphate, and 25‑hydroxyvitamin D may be reasonable, though no formal guidelines exist. Repeat DXA 1–2 years after initiation can detect early bone density loss. If a decline >3–5% at the hip is observed, a change to an alternative SGLT2 inhibitor or addition of a bone‑protective therapy might be warranted. For patients on other SGLT2 inhibitors, routine bone monitoring is not required unless baseline risk factors are present.

Choosing Among SGLT2 Inhibitors

Current evidence suggests that the fracture risk is disproportionately linked to canagliflozin. The FDA has retained the fracture warning in the canagliflozin label (Invokana) but does not include such a warning for dapagliflozin or empagliflozin. Therefore, for patients at increased fracture risk, preference should be given to empagliflozin or dapagliflozin. These agents still provide excellent cardiovascular and renal protection. In patients who are already on canagliflozin without adverse bone events, it may be reasonable to continue with appropriate monitoring. However, de‑novo initiation of canagliflozin in patients with osteoporosis or prior fracture should be avoided.

Combination with Bone‑Protective Therapies

When an SGLT2 inhibitor is needed in a patient with established osteoporosis, concurrent use of antiresorptive therapy (bisphosphonates, denosumab) or anabolic agents (teriparatide) is appropriate. Adequate calcium (1000–1200 mg/day) and vitamin D (600–800 IU/day) intake should be ensured. No drug‑drug interactions have been reported between SGLT2 inhibitors and these agents, and combination does not appear to compromise the diabetes or cardiovascular benefits. Data on the effect of canagliflozin combined with bisphosphonates on BMD are limited, but the rationale is sound.

Broader Cardiovascular and Renal Benefits vs Bone Risk

For the vast majority of patients with type 2 diabetes, the cardiovascular and renal benefits of SGLT2 inhibitors far outweigh any potential bone risk. In the CANVAS Program, despite the increase in fractures, nearly 80% of the cohort had a net benefit in terms of major adverse cardiovascular events, heart failure, and mortality. For patients at low fracture risk—younger, normal BMD, no prior breaks—the bone concern should not dissuade prescription. Even for those with moderate risk, choosing a non‑canagliflozin SGLT2 inhibitor provides the same cardiovascular and renal advantages with a negligible bone signal. The absolute risk of fracture is small (estimated number needed to harm for canagliflozin is around 200 patients over 3 years), whereas the cardiovascular number needed to treat is as low as 50 over 2 years. Thus, the balance is strongly weighted in favor of using the class.

Future Directions and Ongoing Research

Newer SGLT2 inhibitors—sotagliflozin (a dual SGLT1/2 inhibitor) and bexagliflozin—are under evaluation. Preliminary data for sotagliflozin in heart failure and diabetes have not flagged a bone signal, but long‑term fracture data are still maturing. Ongoing large‑scale randomized trials are incorporating fracture adjudication and DXA substudies. The role of SGLT2 inhibitors in patients with type 1 diabetes (both empagliflozin and dapagliflozin have been studied) also requires bone‑specific analysis, as these patients often have lower bone density due to autoimmune issues. Additionally, mechanistic studies are exploring whether the observed changes in bone remodeling are reversible upon discontinuation of the drug. The relation between SGLT2 inhibitor‑induced phosphate changes and FGF‑23 is an active area of investigation, as FGF‑23 itself has been linked to cardiovascular morbidity—potentially tying bone effects to cardiac outcomes.

Guideline Recommendations on SGLT2 Inhibitors and Bone Health

The American Diabetes Association (ADA) Standards of Care Review (2024) note that canagliflozin may increase fracture risk in patients with baseline risk factors and recommend considering bone density evaluation in those patients. The European Society of Cardiology (ESC) guidelines do not differentiate among SGLT2 inhibitors regarding bone risk. The Kidney Disease: Improving Global Outcomes (KDIGO) 2022 guidelines for diabetes in CKD advise that SGLT2 inhibitors are strongly preferred in diabetic kidney disease, but they do not restrict use based on bone status. However, they do note the FDA warning and suggest individualized selection. The National Osteoporosis Foundation has no specific position, but the American Society for Bone and Mineral Research underscores the need for caution when combining multiple osteoporosis risk factors with canagliflozin. Overall, clinical judgment guided by fracture risk assessment remains paramount.

Conclusion

SGLT2 inhibitors are transformative in the management of type 2 diabetes, providing glycemic control along with substantial cardiac and renal protection. The evidence regarding their effect on bone density and fracture risk indicates that the concern is largely limited to canagliflozin, and even for that agent the absolute risk is small. For other members of the class, the bone signal is negligible. Clinicians should stratify fracture risk before prescribing an SGLT2 inhibitor, preferring empagliflozin or dapagliflozin in patients with osteoporosis or prior fracture. Monitoring bone density and mineral metabolism may be prudent in high‑risk patients, especially those on canagliflozin. With these precautions, the benefits of SGLT2 inhibitors overwhelmingly outweigh the potential bone downsides, and the class remains a first‑line choice for most individuals with type 2 diabetes.

For further reading: FDA Safety Communication on canagliflozin, meta‑analysis of fracture risk from Diabetes Care, and the CANVAS Program fracture substudy in Diabetologia.