Effects of SGLT2 inhibition on incident heart failure in carriers of cardiomyopathy-associated genetic variants.
Marston NA, Kany S, Melloni GEM, Jurgens SJ, Kamanu FK, Lai YP, Rämö JT, Raz I, Wiviott SD, Ellinor PT, Sabatine MS, Ruff CT
- DOI
- 10.1038/s41591-026-04439-x
- Record issued
- 2026-08-10
- Engine
- 7.29.0
- Exported
- 2026-09-21
Prepared by Alpha1. This document is confidential: it is intended for the recipient it was shared with and must not be redistributed. The live record at alpha1science.com/verify/b0bd815e-ed54-4cf8-a026-845a008ff135 is authoritative.
How this rating was calculated
- IntegrityIntegrity concern ×2−1★
- ReportingData & code availability not met−0.5★
- ReportingEthical approvals partially met−0.25★
- ReportingKey resources partially met−0.25★
- 01Data and code not shared
The data-availability statement is vague ('data are not available to share'), no sequencing accessions are provided for the exome data, and no analysis code is shared.
“Due to contractual agreements with sponsors, trial data are not available to share.”
Data availabilityFind in source
This Kaimen Rigor review uses Kaimen Rigor reviewers trained on a curated corpus of high-fidelity and retracted papers, with expert supervision and curation. It can still make mistakes; verify each finding against the source before relying on it.
The paper's core strengths are its rigorous randomized, double-blind, placebo-controlled parent trial with blinded adjudication, thorough demographic/biological-variable reporting, detailed statistical methods with assumption checks and plausible arithmetic, and conclusions that are transparently hypothesis-generating. Its main weaknesses are incomplete data/code availability (no sequencing accessions, no custom code), an under-identified investigational product (no dose/formulation), and a missing named regulatory-compliance framework, plus a minor main-text/figure inconsistency (missing minus sign in an ARR CI).
Full-text assessment by two independently sampled reviewer runs of the same model, which converged on six of eight dimension statuses; the one status divergence (data code availability warn vs fail) was resolved by weighing the data-availability statement as adequate, and minor sub-criterion divergences (randomization method detail, exact p-value practice) are noted in the relevant dimensions. Statistical verification recomputed 12 reported tests/effects consistently (threshold-only and exact/permutation p-values out of scope); citation check found 0 retracted and 0 not-found among 30 references; reproducibility links 4/4 live.
Numerical inconsistencies
1 finding · worst lowValues that contradict each other or are impossible for the stated sample: recomputed p-values and test statistics, GRIM/GRIMMER checks on summary numbers, percentages against their own counts, totals against their parts, and estimates against their own confidence intervals.
- Internal contradictions in the reported numbersAssessed
Recomputed 12 tests: 12 consistent, 0 inconsistent; 6 recomputed directly from the reported test statistics, 6 via agent-written checks.
- CONSISTENTreported p < .001 · recomputed p = <.001Recomputed aHR 7.98 (95% CI 3.25–19.59), reported p<0.001
“aHR 7.98, 95% CI 3.25–19.59, P < 0.001”
Taken as given: 3.25–19.59 is a two-sided 95% confidence interval for the aHR of 7.98, not a range, an IQR, or a different interval level; the aHR is a RATIO measure, so the interval is symmetric on the log scale; p<0.001 is the p for THIS estimate, not for another comparison in the same sentenceMethod: back the two-tailed p out of the log-scale CI width and compare it against the printed pHow we recomputed it: pCI(7.98, 3.25, 19.59, 1) - CONSISTENTreported p < .001 · recomputed p = <.001Recomputed aHR 7.89 (95% CI 2.48–25.10), reported p<0.001
“aHR 7.89, 95% CI 2.48–25.10, P < 0.001”
Taken as given: 2.48–25.10 is a two-sided 95% confidence interval for the aHR of 7.89, not a range, an IQR, or a different interval level; the aHR is a RATIO measure, so the interval is symmetric on the log scale; p<0.001 is the p for THIS estimate, not for another comparison in the same sentenceMethod: back the two-tailed p out of the log-scale CI width and compare it against the printed pHow we recomputed it: pCI(7.89, 2.48, 25.1, 1) - CONSISTENTreported p < .001 · recomputed p = <.001Recomputed HR 0.68 (95% CI 0.55–0.83), reported p<0.001
“HR 0.68, 95% CI 0.55–0.83, P < 0.001”
Taken as given: 0.55–0.83 is a two-sided 95% confidence interval for the HR of 0.68, not a range, an IQR, or a different interval level; the HR is a RATIO measure, so the interval is symmetric on the log scale; p<0.001 is the p for THIS estimate, not for another comparison in the same sentenceMethod: back the two-tailed p out of the log-scale CI width and compare it against the printed pHow we recomputed it: pCI(0.68, 0.55, 0.83, 1) - CONSISTENTreported p < .001 · recomputed p = <.001Recomputed HR 8.23 (95% CI 3.61–18.79), reported p<0.001
“HR 8.23, 95% CI 3.61–18.79, P < 0.001”
Taken as given: 3.61–18.79 is a two-sided 95% confidence interval for the HR of 8.23, not a range, an IQR, or a different interval level; the HR is a RATIO measure, so the interval is symmetric on the log scale; p<0.001 is the p for THIS estimate, not for another comparison in the same sentenceMethod: back the two-tailed p out of the log-scale CI width and compare it against the printed pHow we recomputed it: pCI(8.23, 3.61, 18.79, 1) - CONSISTENTreported p = .080 · recomputed p = .081Recomputed HR 2.98 (95% CI 0.88–10.20), reported p=0.08
“HR 2.98, 95% CI 0.88–10.20; P = 0.08”
Taken as given: 0.88–10.20 is a two-sided 95% confidence interval for the HR of 2.98, not a range, an IQR, or a different interval level; the HR is a RATIO measure, so the interval is symmetric on the log scale; p=0.08 is the p for THIS estimate, not for another comparison in the same sentenceMethod: back the two-tailed p out of the log-scale CI width and compare it against the printed pHow we recomputed it: pCI(2.98, 0.88, 10.2, 1) - CONSISTENTreported p = .005 · recomputed p = .006Recomputed HR 0.83 (95% CI 0.73–0.95), reported p=0.005
“HR 0.83; 95% CI 0.73–0.95, P = 0.005”
Taken as given: 0.73–0.95 is a two-sided 95% confidence interval for the HR of 0.83, not a range, an IQR, or a different interval level; the HR is a RATIO measure, so the interval is symmetric on the log scale; p=0.005 is the p for THIS estimate, not for another comparison in the same sentenceMethod: back the two-tailed p out of the log-scale CI width and compare it against the printed pHow we recomputed it: pCI(0.83, 0.73, 0.95, 1) - CONSISTENTreported p = .006 · recomputed p = .006Reviewer 1Two-tailed p for the overall HHF treatment effect HR 0.83 (95% CI 0.73-0.95) from the parent trial primary composite endpoint.
“One of the primary efficacy outcomes of the overall trial was a composite of CV death and HHF, which was significantly reduced by treatment with dapagliflozin compared with placebo (4.9% versus 5.8%, HR 0.83; 95% CI 0.73–0.95, P = 0.005)”
Taken as given: The HR is a ratio-scale estimate and the CI is a two-sided 95% interval.; The p-value is derived from the log-HR and its standard error from the CI width.; Two-tailed test.Method: Recovered the standard error from the 95% CI width of log(HR), z = log(HR)/SE, two-tailed normal p.How we recomputed it: pCI(0.83, 0.73, 0.95, 1) - CONSISTENTreported p = .028 · recomputed p = .028Reviewer 1Two-tailed p for the dapagliflozin vs placebo HHF effect in cardiomyopathy variant carriers HR 0.18 (95% CI 0.04-0.86).
“Among CMP variant carriers, treatment with dapagliflozin resulted in an 82% reduction in incident HHF (HR 0.18, 95% CI 0.04–0.86)”
Taken as given: The HR is a ratio-scale estimate and the CI is a two-sided 95% interval.; The p-value is derived from the log-HR and its standard error from the CI width.; Two-tailed test.Method: Recovered the standard error from the 95% CI width of log(HR), z = log(HR)/SE, two-tailed normal p.How we recomputed it: pCI(0.18, 0.04, 0.86, 1) - CONSISTENTreported p < .001 · recomputed p = <.001Reviewer 1Two-tailed p for the dapagliflozin vs placebo HHF effect in noncarriers HR 0.70 (95% CI 0.57-0.86).
“This was significantly greater than the 30% risk reduction observed in noncarriers (HR 0.70, 95% CI 0.57–0.86, P interaction 0.03)”
Taken as given: The HR is a ratio-scale estimate and the CI is a two-sided 95% interval.; The p-value is derived from the log-HR and its standard error from the CI width.; Two-tailed test.Method: Recovered the standard error from the 95% CI width of log(HR), z = log(HR)/SE, two-tailed normal p.How we recomputed it: pCI(0.70, 0.57, 0.86, 1) - CONSISTENTreported p < .001 · recomputed p = <.001Reviewer 1Two-tailed p for the adjusted HR of HHF in carriers vs noncarriers in the placebo arm aHR 8.06 (95% CI 4.09-15.89).
“translating to an over eightfold increased relative risk among carriers of CMP-associated variants (adjusted hazard ratio (aHR) 8.06, 95% confidence interval (CI) 4.09–15.89, P < 0.001)”
Taken as given: The aHR is a ratio-scale estimate and the CI is a two-sided 95% interval.; The p-value is derived from the log-HR and its standard error from the CI width.; Two-tailed test.Method: Recovered the standard error from the 95% CI width of log(aHR), z = log(aHR)/SE, two-tailed normal p.How we recomputed it: pCI(8.06, 4.09, 15.89, 1) - CONSISTENTreported p = .004 · recomputed p = .002Reviewer 2Baseline chi-squared test of history of HF comparing CMP carriers vs noncarriers
“| History of HF | 22 (18) | 1,235 (10) | 0.004 |”
Taken as given: carriers total 121, so non-HF carriers = 121 − 22 = 99; noncarriers total 12,564, so non-HF noncarriers = 12,564 − 1,235 = 11,329; df = 1 because the table is 2×2; two-sided Pearson chi-squared test as stated in the table footnoteMethod: Pearson chi-squared on 2×2 counts (carrier status × history of HF), two-sidedHow we recomputed it: pChi2x2(22,99,1235,11329) - CONSISTENTreported p = .570 · recomputed p = .505Reviewer 2Baseline chi-squared test of male sex comparing CMP carriers vs noncarriers
“| Male sex | 74 (61) | 8,051 (64) | 0.57 |”
Taken as given: carriers total 121, so female carriers = 121 − 74 = 47; noncarriers total 12,564, so female noncarriers = 12,564 − 8,051 = 4,513; df = 1 because the table is 2×2; two-sided Pearson chi-squared test as stated in the table footnoteMethod: Pearson chi-squared on 2×2 counts (carrier status × sex), two-sidedHow we recomputed it: pChi2x2(74,47,8051,4513)
- lowinternal contradictionMain text reports the absolute risk difference for carriers with prior HF as '17.9% (95% CI 20.6–56.5%)', but this CI does not bracket the point estimate 17.9%; Fig. 3b gives the same statistic as '17.9% (95% CI −20.6% to 56.5%)', indicating a missing minus sign in the main text.
“The absolute risk differences were 4.1% (95% CI 0.3–7.8%) in noncarriers and 17.9% (95% CI 20.6–56.5%) in carriers”
Main, 'Effect of dapagliflozin by histo…Find in source
Overstated conclusions
1 finding · worst lowConclusions that reach past what the paper's own results support — including a significance claim that no longer holds when the statistic is recomputed, and efficacy resting on an unvalidated surrogate endpoint.
- Conclusions only partially backed by the presented evidenceAssessed
8 major claims checked against the paper's own evidence: all adequately supported.
- partialReviewer 1The findings raise the possibility that SGLT2 inhibitor treatment should be started early to prevent HF in carriers of P/LP cardiomyopathy variants.The data support a greater relative and absolute benefit in carriers, but the claim is appropriately framed as hypothesis-generating given the post-hoc analysis and small subgroup, and the authors themselves call for prospective confirmation.Evidence: HR 0.18 in carriers, 13.0% ARR, and a pronounced effect in carriers without prior HF (no events in the dapagliflozin arm).
“The findings from this cohort of older and high-risk patients raise the possibility that SGLT2 inhibitor treatment should be started early to prevent HF in individuals who carry P/LP cardiomyopathy variants.”
AbstractFind in source - partialReviewer 2In carriers without prior HF, dapagliflozin reduced the absolute risk of HHF by 12.8%, with no events in the dapagliflozin arm.The reported ARR and interaction are as stated, but the zero-event dapagliflozin arm means the HR could not be calculated and the estimate rests on a very small number of events, making it unstable.Evidence: 12.8% ARR in carriers vs 0.6% in noncarriers, P interaction 0.01; 'none of the patients allocated to dapagliflozin experiencing incident HF'
“This 12.8% ARR was significantly greater than what was observed in noncarriers without HF (ARR 12.8% (95% CI 3.1–22.4%) versus 0.6% (95% CI 0.1–1.1%), P interaction 0.01)”
Main, 'Effect of dapagliflozin by histo…Find in source - partialReviewer 2CMP variants are actionable genotypes that identify patients who derive the greatest benefit from SGLT2 therapy, supporting early preventive treatment.The interaction supports greater benefit in carriers, but labeling variants 'actionable genotypes' and implying early preventive treatment reaches beyond the post-hoc subgroup evidence, which the authors themselves describe as hypothesis-generating.Evidence: Interaction P 0.03 and larger ARR in carriers; the paper itself calls the findings 'hypothesis-generating' and requires confirmation in 'a prospective, dedicated trial'
“we demonstrate that CMP variants are actionable genotypes, in which therapeutic management may be influenced by the results of genetic testing”
DiscussionFind in source - supportedReviewer 1Dapagliflozin reduced HHF more strongly in CMP-variant carriers (HR 0.18) than in noncarriers (HR 0.70), with a significant interaction (P=0.03).The reported HRs, CIs, and interaction p-value are presented in the results and are internally consistent.Evidence: HR 0.18 (0.04–0.86) in carriers vs HR 0.70 (0.57–0.86) in noncarriers, P interaction 0.03 (Fig. 2).
“dapagliflozin lowered the risk of HHF more strongly in carriers (hazard ratio 0.18, 95% confidence interval 0.04–0.86) than in noncarriers (hazard ratio 0.70, 95% confidence interval 0.57–0.86; P interaction 0.03)”
AbstractFind in source - supportedReviewer 1Absolute risk reduction was 13.0% in carriers vs 1.0% in noncarriers (P interaction 0.03).The ARR values are internally consistent with the reported event proportions (16.1% to 3.1% = 13.0%) and the NNT of 7.7 equals 1/0.13.Evidence: Kaplan–Meier rates 16% to 3.6% (text: 16.1% to 3.1%); ARR 13.0% (2.4–23.6%) vs 1.0% (0.4–1.6%); NNT 7.7.
“This 13.0% (2.4–23.6%) absolute risk reduction (ARR) was significantly greater than the 1.0% (0.4–1.6%) ARR observed in noncarriers ( P interaction 0.03) and translated to a number needed to treat of just 7.7”
ResultsFind in source - supportedReviewer 1Carriers of CMP-associated variants have an approximately eightfold increased risk of HHF in the placebo arm.The adjusted HR of 8.06 (95% CI 4.09–15.89) supports an eightfold elevation, though the unadjusted incidence ratio (16% vs 3.5%) is about 4.6-fold; the claim is anchored to the adjusted estimate and is therefore supported.Evidence: aHR 8.06 (95% CI 4.09–15.89, P<0.001).
“translating to an over eightfold increased relative risk among carriers of CMP-associated variants (adjusted hazard ratio (aHR) 8.06, 95% confidence interval (CI) 4.09–15.89, P < 0.001)”
ResultsFind in source - supportedReviewer 2Dapagliflozin reduces the risk of hospitalization for heart failure more strongly in carriers of cardiomyopathy-associated variants than in noncarriers.The presented interaction term and the two hazard ratios with confidence intervals directly support a greater relative benefit in carriers.Evidence: HR 0.18 (95% CI 0.04–0.86) in carriers vs HR 0.70 (95% CI 0.57–0.86) in noncarriers, P interaction 0.03 (Fig. 2)
“dapagliflozin lowered the risk of HHF more strongly in carriers (hazard ratio 0.18, 95% confidence interval 0.04–0.86) than in noncarriers (hazard ratio 0.70, 95% confidence interval 0.57–0.86; P interaction 0.03)”
AbstractFind in source - supportedReviewer 2Carriers of CMP-associated variants have markedly (over eightfold) increased risk of incident heart failure hospitalization.The adjusted hazard ratio and confidence interval, plus consistency across DCM, TTN, and HCM subgroups, support the claim.Evidence: aHR 8.06 (95% CI 4.09–15.89), P < 0.001 in the placebo arm; consistent across DCM (7.98), TTN (10.72), and HCM (7.89)
“translating to an over eightfold increased relative risk among carriers of CMP-associated variants (adjusted hazard ratio (aHR) 8.06, 95% confidence interval (CI) 4.09–15.89, P < 0.001)”
Main, 'Association of carrier status wi…Find in source
Efficacy claim is anchored to an adequate endpoint and a meaningful effect.
- ADEQUATESurrogate endpointThe primary endpoint is hospitalization for heart failure, a hard clinical outcome.
“treatment with dapagliflozin resulted in an 82% reduction in incident HHF (HR 0.18, 95% CI 0.04–0.86)”
- ADEQUATEEffect sizeThe effect size is large: 82% relative risk reduction, 13% absolute risk reduction, NNT 7.7, which is clinically meaningful.
“This 13.0% (2.4–23.6%) absolute risk reduction (ARR) was significantly greater than the 1.0% (0.4–1.6%) ARR observed in noncarriers”
Data authenticity concerns
1 finding · worst lowAn adversarial read for patterns associated with data that may not be genuine: results that look too clean, implausibly large effects, duplicated data or images, and methods that do not match the results reported.
- Implausibly large reported effectAssessed
2 integrity concerns flagged (0 high).
- lowimplausible effectThe treatment-effect interaction rests on a small number of events in the carrier subgroup (9 placebo events in 56 carriers), producing a very wide HR CI (0.04–0.86) and a borderline interaction p (0.03); the authors transparently acknowledge the small subgroup and post-hoc nature, so this is a hypothesis-generating finding rather than a contradiction.
“dapagliflozin lowered the risk of HHF more strongly in carriers (hazard ratio 0.18, 95% confidence interval 0.04–0.86) than in noncarriers (hazard ratio 0.70, 95% confidence interval 0.57–0.86; P interaction 0.03)”
AbstractFind in source
Reporting gaps
3 findings · worst highRequired detail the manuscript never states — study design, biological variables, ethics approval and consent, key resources, statistical reporting, data and code availability, and overall transparency.
- Data and code not sharedAssessed
- Ethics/consent reporting incompleteAssessed
- Key resources under-identified (antibodies, cell lines, RRIDs)Assessed
Prior work on SGLT2 inhibitors reducing HHF, the heritability of CMPs, and the prevalence/risk of P/LP variants is cited with specific cohorts (Spanish DCM cohort, UK Biobank). The rationale links the premise to the objective (testing dapagliflozin benefit in variant carriers) and the hypothesis follows. Limitations of the design are extensively addressed in the Discussion (small carrier number, post-hoc nature, absent imaging, generalizability).
“Sodium–glucose cotransporter 2 (SGLT2) inhibitors such as dapagliflozin have been shown to prevent hospitalization for HF (HHF) across the spectrum of left ventricular ejection fraction (LVEF)”
“Whether SGLT2 inhibitor treatment is of benefit in asymptomatic individuals with a CMP variant is not known.”
“Whether SGLT2 inhibitor treatment is of benefit in asymptomatic individuals with a CMP variant is not known.”
“Apart from increased surveillance and monitoring, the therapeutic consequence of identifying CMP variants, particularly in asymptomatic individuals, remains unclear”
Randomization came from the parent double-blind, randomized, placebo-controlled trial; blinding is described (double-blind, and events adjudicated by a committee blinded to treatment assignment). Inclusion/exclusion for the genetic substudy (consent + available WES, stringent variant curation) is detailed. Outlier/missing-data handling is stated (available-case, no imputation needed). No a priori power/sample-size calculation is reported, which is typical for a post-hoc subgroup analysis. Independent replication is appropriately not claimed and a prospective trial is stated as needed.
“This is a secondary analysis of DECLARE-TIMI 58, a double-blind, randomized, placebo-controlled trial”
“All events were adjudicated by a central clinical events committee blinded to treatment assignment”
“This analysis was not prespecified.”
“This is a secondary analysis of DECLARE-TIMI 58, a double-blind, randomized, placebo-controlled trial”
“All events were adjudicated by a central clinical events committee blinded to treatment assignment”
“Analyses were performed on an available-case basis; no imputation methods were required due to minimal missing data.”
Both sexes were enrolled (male 61% carriers vs 64% noncarriers), so no single-sex justification is required. Age (63.4±6.4), BMI, comorbidities, NT-proBNP, and hs-TnT are reported. Human demographics including ethnicity are provided in Table 1 and Extended Data Tables.
“Male sex | 74 (61) | 8,051 (64) | 0.57”
“Male sex | 74 (61) | 8,051 (64)”
“| Male sex | 74 (61) | 8,051 (64) | 0.57 |”
“| Age | 63.4 ± 6.4 | 64 ± 6.9 | 0.43 |”
“| Ethnicity | Asian 7 (6) | Asian 1,140 (9) | 0.01 |”
The paper states 'The trial protocol was approved by the institutional review board at each participating site, and all participants provided written informed consent,' which adequately covers approval and consent. However, no named regulatory framework (Declaration of Helsinki, ICH-GCP) is explicitly cited, so regulatory_compliance is not reported. Because all applicable ethical elements must be adequate for a pass, this gap yields a warn.
“The trial protocol was approved by the institutional review board at each participating site, and all participants provided written informed consent.”
“The trial protocol was approved by the institutional review board at each participating site, and all participants provided written informed consent.”
Software tools are thoroughly identified (R 4.4.2, BWA 0.7.15-r1140, GATK 4.1.4.1, VEP v105, LOFTEE, with GitHub URLs). The investigational product 'dapagliflozin' is named along with the sponsor (AstraZeneca), but no dose, regimen, or formulation is given in the text, so reagents_identified is reported_but_inadequate. Antibodies, cell lines, mycoplasma, and organisms are not applicable to this clinical trial.
“All analyses were performed using R version 4.4.2 (R Foundation).”
“dapagliflozin treatment”
“All analyses were performed using R version 4.4.2 (R Foundation).”
“aligned to GRCh38 with BWA 0.7.15-r1140 and converted to SAM specification 1.6 CRAM files with GATK 4.1.4.1”
Cox proportional-hazard models, Wilcoxon rank-sum, chi-squared, Wald tests, and Schoenfeld residuals are named; the proportional-hazards assumption was tested with no violations. HRs and ARRs are reported with 95% CIs throughout. Software (R 4.4.2) is identified. Data presentation uses KM curves with per-group n and CIs. Exact p-values are given for most comparisons (e.g., P interaction 0.03, 0.01), with thresholds (P < 0.001) used only for very small values. Checkable proportions are plausible (221/6,291 = 3.5%; 9/56 = 16%; NNT 1/0.13 = 7.7).
“We tested the proportional hazards assumption using Schoenfeld residuals and found no violations.”
“dapagliflozin lowered the risk of HHF more strongly in carriers (hazard ratio 0.18, 95% confidence interval 0.04–0.86) than in noncarriers (hazard ratio 0.70, 95% confidence interval 0.57–0.86; P interaction 0.03)”
“the incidence of HHF was 3.5% (221/6,291) in noncarriers and 16% (9/56) in carriers”
“We tested the proportional hazards assumption using Schoenfeld residuals and found no violations.”
“Among CMP variant carriers, treatment with dapagliflozin resulted in an 82% reduction in incident HHF (HR 0.18, 95% CI 0.04–0.86).”
“the incidence of HHF was 3.5% (221/6,291) in noncarriers and 16% (9/56) in carriers”
A data-availability statement exists but offers no concrete access route, platform, or managed-access mechanism, so it is reported_but_inadequate. The study generated non-identifiable sequencing data (13,184 exomes) for which accessions would be expected but none are given. No bespoke analysis code is shared. None of the applicable availability criteria is adequate, so the dimension fails.
“Due to contractual agreements with sponsors, trial data are not available to share.”
“Due to contractual agreements with sponsors, trial data are not available to share.”
Methods are sufficiently detailed for replication (sequencing QC, variant curation, models). DECLARE-TIMI 58 registration (NCT01730534) is given. Secondary exploratory endpoints (CV death, composite, all-cause mortality) are reported as non-significant. The limitations section is extensive and conclusions are appropriately hedged ('hypothesis-generating', 'need to be confirmed in a prospective, dedicated trial'). Funding and competing interests are detailed. The reporting guideline is referenced only via a linked 'Reporting summary' rather than an explicitly named checklist, which is a minor gap that does not prevent a pass.
“These results need to be confirmed in a prospective, dedicated trial of preventive HF treatments in carriers of P/LP cardiomyopathy-associated variants.”
“Further information on research design is available in the linked to this article.”
“These results need to be confirmed in a prospective, dedicated trial of preventive HF treatments”
“Another important limitation of this analysis is the absence of systematic cardiac imaging.”
Registered (2 IDs: ClinicalTrials.gov). No reporting guideline cited.
Broken references and links
None foundReferences checked against Crossref, OpenAlex and Retraction Watch for retractions and resolvability, plus declared data and code links probed for whether they resolve to content matching the paper.
Checked — nothing surfaced.
Checked 30 references by DOI: 5 verified — 25 no DOI (shown, not verified).
- NO DOIEpidemiology and aetiology of heart failure.No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIDapagliflozin and cardiovascular outcomes in type 2 diabetes.No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIDapagliflozin in patients with heart failure and reduced ejection fraction.No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIDapagliflozin in heart failure with mildly reduced or preserved ejection fraction.No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOILong-term outcomes of dilated cardiomyopathy diagnosed during childhood: results from a national population-based study of childhood cardiomyopathy.No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOI2023 ESC Guidelines for the management of cardiomyopathies: developed by the task force on the management of cardiomyopathies of the European Society of Cardiology (ESC).No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOI2022 AHA/ACC/HFSA Guideline for the Management of Heart Failure: a report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines.No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIPrevalence and disease expression of pathogenic and likely pathogenic variants associated with inherited cardiomyopathies in the general population.No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIAssessing the role of rare genetic variation in patients with heart failure.No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOICardiomyopathy-associated gene variants in atrial fibrillation.No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIPathogenic cardiomyopathy-associated gene variants and prognosis in atrial fibrillation: results in 18,000 clinical trial participants.No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIAssociation of genetic variants with outcomes in patients with nonischemic dilated cardiomyopathy.No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIPredicted deleterious variants in cardiomyopathy genes prognosticate mortality and composite outcomes in the UK Biobank.No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIGenomics-first evaluation of heart disease associated with Titin-truncating variants.No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIPenetrance of dilated cardiomyopathy in genotype-positive relatives.No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIThe ensembl variant effect predictor.No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIThe mutational constraint spectrum quantified from variation in 141,456 humans.No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIClinVar: improving access to variant interpretations and supporting evidence.No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIEvidence-based assessment of genes in dilated cardiomyopathy.No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIGenetic evaluation of cardiomyopathy: a clinical practice resource of the American College of Medical Genetics and Genomics (ACMG).No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIInternational evidence based reappraisal of genes associated with arrhythmogenic right ventricular cardiomyopathy using the Clinical Genome Resource framework.No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIRare and common genetic variation underlying the risk of hypertrophic cardiomyopathy in a national biobank.No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIMonogenic and polygenic contributions to atrial fibrillation risk: results from a national biobank.No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIIntegrated allelic, transcriptional, and phenomic dissection of the cardiac effects of titin truncations in health and disease.No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIEffect of dapagliflozin on heart failure and mortality in type 2 diabetes mellitus.No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
4 data/code links checked; 4 live.
- datahttps://clinicaltrials.gov/ct2/show/NCT01730534LIVEHTTP 200Resolves, but the content could not be matched to the paper.
- datahttps://clinicaltrials.gov/ct2/show/NCT06677060LIVEHTTP 200Resolves, but the content could not be matched to the paper.
- codeGitHubLIVEHTTP 200https://github.com/gatk-workflows/gatk4-exome-analysis-pipelineResolves to GitHub (code repository).
- codeGitHubLIVEHTTP 200https://github.com/konradjk/lofteeResolves to GitHub (code repository).
Copyediting
1 finding · worst lowWording, consistency and formatting errors that need correcting before submission.
- Wording or formatting errors that need correctingAssessed
4 copyedit issues flagged (2 major): mostly consistency.
- MAJORconsistencyResults, Effect of dapagliflozin by history of HF“The absolute risk differences were 4.1% (95% CI 0.3–7.8%) in noncarriers and 17.9% (95% CI 20.6–56.5%) in carriers”→ Change to '17.9% (95% CI −20.6% to 56.5%)' to match the Fig. 3b caption.The minus sign on the lower bound of the ARR CI is missing in the main text but present in the figure caption.
- MAJORconsistencyMain, 'Effect of dapagliflozin by history of HF'“17.9% (95% CI 20.6–56.5%) in carriers”→ Change to '17.9% (95% CI −20.6% to 56.5%)' to match Fig. 3b and so the CI brackets the point estimate.A missing minus sign makes the CI appear not to contain the point estimate; Fig. 3b shows the correct value.
- MINORconsistencyAssociated Data / Data Availability“Due to contractual agreements with sponsors, trial data are not available to share.”→ State the data-availability statement once in a single location.The identical data-availability sentence appears twice (in 'Data availability' and in 'Associated Data').
- MINORconsistencyMain vs Table 1“NT-proBNP levels were also higher in carriers ... (P < 0.001)”→ Reconcile the NT-proBNP p-value between text ('P < 0.001') and Table 1 ('0.0004').Both values are consistent with strong significance, but a single value should be used.
In a post-publication audit, the published work is methodologically sound in its core randomized comparison and self-consciously hypothesis-generating; an informed reader should weigh the small carrier subgroup (9 placebo events; wide HR CI 0.04–0.86; interaction p=0.03), the absence of sequencing data accessions and analysis code, and the main-text ARR CI typo (missing minus sign), which is a minor internal inconsistency that would warrant a small erratum rather than re-analysis. No recomputed statistical error was found, and the headline interaction finding is appropriately hedged.
- 1.HIGHcopyeditCorrect the missing minus sign in the Results text for carriers with prior HF — change '17.9% (95% CI 20.6–56.5%)' to '17.9% (95% CI −20.6% to 56.5%)' to match Fig. 3b and so the CI brackets the point estimate.This is a demonstrable internal contradiction between the main text and figure caption (also flagged by the integrity lens) that an informed reader will notice and which warrants an erratum.
- 2.HIGHdata codeProvide a concrete access route for the whole-exome sequencing data (13,184 exomes) — e.g., deposit in a controlled-access repository (dbGaP/EGA) with accession numbers, or name a managed-access committee with request conditions — in the Data availability section.Generated sequencing data are a key reproducibility resource; the absence of any accession leaves the substudy's core data effectively unavailable to readers.
- 3.HIGHdata codeDeposit the study-specific analysis code (variant curation and the Cox/pseudo-value modeling) in a version-controlled public repository with a persistent identifier and cite it in Methods, Statistical analysis.Only third-party pipeline URLs (GATK, LOFTEE) are cited; the custom analysis code is not shared, limiting independent re-analysis of the subgroup interaction.
- 4.MEDIUMethicsAdd an explicit regulatory-compliance statement (e.g., 'conducted in accordance with the Declaration of Helsinki' / ICH-GCP) to the Methods ethics paragraph alongside the IRB approval statement.IRB approval and consent are stated but no named compliance framework is given, which is the sole reason the ethics dimension does not pass.
- 5.MEDIUMrigorState the dapagliflozin dose, regimen, formulation, and manufacturer in Methods, Study design (the parent DECLARE-TIMI 58 dose was 10 mg).The investigational product is the key resource in a drug trial and is currently named without dose/formulation, so the intervention cannot be fully reproduced from the text.
- 6.MEDIUMreportingCite a named reporting guideline (e.g., STROBE for the observational genetic analysis, or CONSORT for the parent trial) instead of only the generic linked 'Reporting summary'.A named checklist is the current reporting-guideline gap in an otherwise transparent methods section.
- 7.MEDIUMcopyeditReconcile the NT-proBNP p-value between the main text ('P < 0.001') and Table 1 ('0.0004') so a single value is used.Both values indicate strong significance, but the inconsistency is a minor internal-consistency issue a reader or editor will catch.
- 8.MEDIUMrigorKeep the small-event basis of the carrier-subgroup finding prominent: the treatment-effect interaction rests on 9 placebo events in 56 carriers (HR 0.18, 95% CI 0.04–0.86; interaction p=0.03) and should remain framed as hypothesis-generating.The wide CI and borderline interaction p make this a hypothesis-generating signal, not a demonstration of benefit; the authors already acknowledge this and it should stay foregrounded.
- 9.LOWstatisticsWhere feasible, report exact p-values instead of thresholds (e.g., the overall HHF HR and the aHR 8.06 reported as 'P < 0.001').Threshold-only reporting of a computed p-value is imprecise; exact values improve precision while remaining consistent with the current conclusions.
- 10.LOWcopyeditRemove the duplicated data-availability sentence that appears in both 'Data availability' and 'Associated Data' so the statement appears once.The identical sentence is stated twice, a minor redundancy in the published layout.
The star rating is the report’s one-glance summary. Every paper starts at 5★ and loses stars for the concrete problems the review finds — so a rating is never a vague average, it’s a running total you can read line by line under “How this rating was calculated.”
- Reporting — 8 dimensionseach dimension that fully fails−½★
- each dimension partially met−¼★
- Statistics · Integrity · Claimseach serious problem−1★
- each medium problem−½★
- Citationseach retracted or unverifiable reference−¼★
- Copyeditonly when the manuscript needs a full edit−½★
The rating never drops below 1★, and a demonstrable critical failure (an impossible statistic, a proven ethics violation) caps it at 1★ on its own — so the stars can never look healthy when the verdict is CRITICAL.
The rating draws on a panel of agents. Three independent Kaimen Rigor reviewers grade the eight dimensions below across several independent passes (the shown verdict is their majority vote — steadier than any single run), isolate the paper’s major claims and check its own evidence backs them, and flag integrity concerns. Alongside them, a citation agent resolves every reference against Crossref, OpenAlex, and Retraction Watch; a statistics agent recomputes reported tests; and rule-based checks verify that declared data/code links actually resolve. Full text is required — an abstract-only submission is not analyzed.
Graded against NIH, MDAR, ARRIVE 2.0, CONSORT, EQUATOR, and RRID guidelines. A dimension that doesn’t apply to the study type is skipped, never penalized.
This Kaimen Rigor review is model-assisted and is not a substitute for formal expert review. It complements human evaluation by surfacing potential methodological concerns — verify each finding against the source.