Interpretation of EMPA-ELDERLY Trial Results on Muscle Mass Loss
There was no statistically significant difference in muscle mass loss with empagliflozin versus placebo because the CI crossed 0.
Understanding the Statistical Interpretation
The EMPA-ELDERLY trial examined the effects of empagliflozin 10 mg versus placebo in older adults with type 2 diabetes mellitus, with muscle mass change as a secondary outcome. The reported placebo-adjusted mean change in muscle mass at 52 weeks was -0.61 kg (95% CI, -1.61 to 0.39).
This result should be interpreted as follows:
- When a 95% confidence interval crosses zero (includes both negative and positive values), the result is not statistically significant at the p<0.05 level
- In this case, the confidence interval (-1.61 to 0.39) includes zero, indicating no statistically significant difference between empagliflozin and placebo for muscle mass loss
- The point estimate (-0.61 kg) suggests a trend toward greater muscle mass loss with empagliflozin, but this did not reach statistical significance
Clinical Significance vs. Statistical Significance
While there was no statistically significant difference in muscle mass loss, it's important to consider:
- The point estimate (-0.61 kg) represents approximately a 2% loss in muscle mass for an average older adult
- This magnitude of change may not be clinically significant in most patients
- The EMPA-ELDERLY trial specifically reported that empagliflozin "improved glucose control and reduced body weight without compromising muscle mass or strength in elderly adults with T2DM" 1
Cardiovascular Benefits of Empagliflozin
Despite the non-significant finding regarding muscle mass, empagliflozin has demonstrated important cardiovascular benefits:
- In the EMPA-REG OUTCOME trial, empagliflozin reduced major adverse cardiovascular events by 14% (HR 0.86; 95% CI 0.74-0.99) 2, 3
- Cardiovascular death was reduced by 38% (HR 0.62; 95% CI 0.49-0.77) 2, 3
- Hospitalization for heart failure was reduced by 35% (HR 0.65; 95% CI 0.50-0.85) 2
Common Pitfalls in Statistical Interpretation
When interpreting confidence intervals:
- For hazard ratios, risk ratios, or odds ratios: CI crossing 1.0 indicates no statistical significance
- For absolute differences (as in this case): CI crossing 0 indicates no statistical significance
- Confusing these two different statistical contexts is a common error
Clinical Implications
For elderly patients with type 2 diabetes:
- Empagliflozin provides significant cardiovascular and glycemic benefits
- The non-significant trend toward muscle mass loss should be monitored but does not outweigh the proven benefits
- The European Society of Cardiology recommends SGLT2 inhibitors with proven CV benefit in patients with T2DM and cardiovascular disease 2
- The American Diabetes Association recommends SGLT2 inhibitors as preferred agents for patients with established ASCVD, heart failure, or CKD 2, 4
In conclusion, while there was a numerical difference in muscle mass loss with empagliflozin versus placebo in the EMPA-ELDERLY trial, this difference was not statistically significant as evidenced by the confidence interval crossing zero.