Reviews/SGLT2 Inhibitors/SGLT2 inhibitors do not significantly reduce CV mortality in HFpEF
· SG-HP-02 · v2
The question is whether SGLT2 inhibitors cut deaths from cardiovascular causes specifically in heart failure patients whose pumping function is preserved. Across the studies that count here, none measures that exact outcome in a large, well-powered way. The one study that reports cardiovascular deaths directly [12] found no difference, but it followed only 52 patients, far too few to settle the matter. The major trial evidence [15][16] bundles cardiovascular death together with heart-failure hospitalisations into a single combined figure, so the death-only part cannot be pulled out. A large real-world cohort [13] shows the benefit falls on keeping people out of hospital rather than on hard events like heart attack or stroke. The panel landed where it did because the available evidence leans the claim's way without ever pinning the death outcome on its own with enough statistical weight.
No included primary study demonstrates a significant standalone reduction in cardiovascular mortality in HFpEF. The single-centre retrospective series [12] reports no significant difference in cardiovascular or all-cause mortality, but with only 52 patients (25 versus 27) it is severely underpowered. The EMPEROR-Preserved analyses [15][16] report only the composite of cardiovascular death or heart-failure hospitalisation (hazard ratio 0.83), never death alone. The Medicare cohort [13] shows the SGLT2i advantage concentrated in heart-failure hospitalisation (hazard ratio 0.65 versus DPP4i) with no significant effect on myocardial infarction or stroke (hazard ratio 0.90, 95% CI 0.79-1.02). The pattern is directionally consistent with the claim but does not isolate the mortality endpoint with adequate power. The contested point: The division is whether the absence of a demonstrated death reduction reflects a true null or simply the lack of an adequately powered, death-isolating analysis in the included set. The composite hazard ratio of 0.83 in [15][16] contains cardiovascular death, and the corpus text cannot decompose it, so a mortality contribution can be neither confirmed nor excluded from this evidence.
All 21 papers behind this claim with their DOIs and what each does to it, the panel’s arbitration, the evidence gap map and the open research questions. Free, and sent to your inbox.
We’ll also send you other claims from SGLT2 Inhibitors, one at a time, and tell you when a release or a pivotal trial changes a grade. Unsubscribe by replying.
One reviewer's only job is to attack the finding. The challenges below were weighed before the grade was set.
The only included primary study reporting CV mortality directly [12] has N=52 and no significant CV-mortality difference is absence of evidence, not evidence of absence; it is grossly underpowered to establish the null the claim asserts.
The landmark HFpEF RCT evidence [15][16] reports the COMPOSITE of CV death or HF hospitalization (HR 0.83), not isolated CV mortality; the corpus text cannot decompose it, so a CV-mortality contribution cannot be excluded from the included evidence.
Where it remains uncertain. The one direct study is tiny (N=52) [12] and the composite hazard ratio of 0.83 [15][16] cannot be separated into its death component, so a true null and an undetected benefit remain hard to distinguish.
Aleatory
With only 52 patients in [12], chance variation alone could mask or mimic a mortality difference; the null result carries wide uncertainty.
Epistemic
[15] and [16] report cardiovascular death or heart-failure hospitalisation together (hazard ratio 0.83); the death-only component is unmeasured in the included text and could be resolved with the underlying trial data.
Inferential
[13] compares SGLT2i against DPP4i and GLP-1RA, not placebo, with residual confounding acknowledged, so inferring a placebo-relative mortality effect from it is indirect.
The load-bearing papers from a corpus of 21, each classified by what it does to the claim. Every DOI is taken from the corpus the claim was graded on.
Only included primary study reporting CV mortality directly in HFpEF; finds no significant difference, but N=52 makes this uninformative about a true null
“No statistically significant differences were found in clinical outcomes, including 30- and 90-day HF hospitalizations, all-cause readmissions, cardiovascular mortality, or all-cause mortality.”
Panel auditInternal validity: low · scope of inference: Effect of inpatient SGLT2i initiation on prescription rates and short-term outcomes in HFpEF.Key limitations: n=52 underpowered · single-center retrospective · short follow-up
Observational HFpEF cohort shows benefit concentrated in HF hospitalization and not in hard atherothrombotic events; consistent with mortality not being the driver
“Among HFpEF patients, initiation of SGLT2i versus DPP4i (cohort 2a; n = 17,493) was associated with lower risk of HHF (HR: 0.65 [0.61, 0.69]) but not MI or stroke (HR: 0.90 [0.79, 1.02])”
Panel auditInternal validity: low · scope of inference: SGLT2i versus DPP4i/GLP-1RA HHF and MI/stroke outcomes in HFpEF.Key limitations: no isolated CV-mortality estimate · active comparators not placebo · residual confounding
Landmark HFpEF RCT reports a significant COMPOSITE of CV death or HHF, not isolated CV mortality; corpus cannot decompose the CV-death contribution
“In patients with LVEF ≥ 50%, empagliflozin reduced the risk of cardiovascular death or HHF (the primary endpoint) by 17% versus placebo (hazard ratio (HR) 0.83; 95% confidence interval (CI): 0.71-0.98, P = 0.024).”
Panel auditInternal validity: high · scope of inference: Empagliflozin effect on the CV death or HHF composite across LVEF strata in HFpEF.Key limitations: composite endpoint not CV mortality · CV mortality not isolated in abstract
3 of 21 papers shown. The full review lists all of them, with the panel’s arbitration →
Claims are graded at each quarterly release, and re-graded between releases only when a pivotal trial lands. Every change is a new version of the claim.
The surviving primary research is consistent with the null but too thin and indirect on the specific CV-mortality endpoint to carry a confident verdict.
Release 2026-q4 re-graded every claim from its corpus under protocol v0.19. An earlier release's grade is shown for the record; it came from an earlier protocol and is not a movement.
Assumptions that carry weight but are not themselves established.
The panel records the gap that would most move the claim, and the study that would close it.
Isolated cardiovascular-mortality endpoint in a powered HFpEF trial. Every included primary source either reports a composite [15][16] or is underpowered [12], so the specific endpoint the claim names is never tested on its own with adequate numbers. The panel rates this the highest-importance open gap for the claim, and names the study design that would close it: A large placebo-controlled HFpEF trial powered for cardiovascular death as a pre-specified standalone endpoint, or a component-level decomposition of EMPEROR-Preserved reporting cardiovascular death alone.
Open question: Does the EMPEROR-Preserved composite benefit (hazard ratio 0.83) include any significant reduction in cardiovascular death on its own? Why it is still open: [15] and [16] report only the composite of cardiovascular death or heart-failure hospitalisation and the corpus text cannot decompose it.
Priors is a sceptical first reader, not peer review, and not medical advice.
Release 2026-q4 · graded 23 Oct 2026 · protocol v0.19