Chronic kidney disease and heart failure disproportionately affect populations of African ancestry. Prior Mendelian randomisation (MR) studies examining the causal relation between kidney function and heart failure subtypes have been conducted exclusively in European ancestry populations. This preprint reports a forward two-sample MR analysis designed to evaluate whether genetically predicted estimated glomerular filtration rate (eGFR) causally influences risk of heart failure with preserved ejection fraction (HFpEF) and heart failure with reduced ejection fraction (HFrEF) in individuals of African ancestry.
The study aims to provide ancestry-matched genetic evidence to address potential biases arising from cross-ancestry instrument/outcome mismatches and to inform whether genetically determined kidney function contributes to subtype-specific heart failure risk in this population.
Genetic instruments for eGFR were selected from an African ancestry genome-wide association study (GWAS) comprising N = 67,943 participants. Instrument selection used genome-wide significance thresholds and linkage disequilibrium clumping with an African ancestry reference panel to retain independent variants suitable for MR.
After these procedures, six independent single-nucleotide polymorphisms (SNPs) were retained as instruments. Instrument strength metrics reported included F-statistics ranging from 30.5 to 107.3, and the combined instruments explained R² = 0.62% of eGFR variance in the GWAS sample.
Outcome summary statistics for heart failure subtypes were sourced from the Million Veteran Program. Case-control counts for each subtype were:
These ancestry-matched outcome data were used in two-sample MR analyses with the African ancestry eGFR instruments.
The authors performed a forward two-sample MR framework. The primary analytic approach was the inverse-variance weighted (IVW) MR estimate combining effects from the six independent eGFR-associated SNPs to obtain an overall causal estimate of genetically predicted eGFR on each heart failure subtype.
Effect estimates are presented as odds ratios (OR) per genetically predicted unit change in eGFR, with 95% confidence intervals and p values reported for the primary IVW analyses.
Sensitivity analyses were conducted to probe robustness of the primary IVW findings. The report states sensitivity results were directionally consistent with the primary analysis. Tests for heterogeneity between instrument-specific estimates and for directional pleiotropy were performed; no evidence of heterogeneity or directional pleiotropy was observed in the summary provided.
Specific sensitivity methods (for example, MR-Egger, weighted median) and their numerical outputs were not detailed in the abstract; the authors report directional consistency and negative findings for heterogeneity and pleiotropy.
Power considerations were explicitly addressed. The report provides minimum detectable effect sizes at 80% statistical power given the instrument strength and sample sizes: an OR of 1.28 for HFpEF and an OR of 1.22 for HFrEF. These thresholds indicate the study was powered to detect only relatively large causal effects and had limited ability to detect smaller effect sizes.
The primary IVW MR results were null for both heart failure subtypes:
Sensitivity analyses did not materially change the directional interpretation. The authors report no evidence of heterogeneity or directional pleiotropy across instruments.
The analyses did not find evidence that genetically predicted eGFR causally affects risk of HFpEF or HFrEF in African ancestry samples given the data and instruments used. However, the authors caution that these null findings should be considered inconclusive due to current power limitations: the instruments explained a small proportion of eGFR variance (R² = 0.62%) and the MR analyses were only powered to detect moderately large odds ratios (≥1.22–1.28).
Additional limitations implied by the reported results include the small number of retained instruments (six SNPs) and the absence of detailed sensitivity-method outputs in the abstract. The study is a preprint and has not yet been peer reviewed, which is important when interpreting findings and considering clinical implications.
The authors recommend larger ancestry-matched genetic studies to increase instrument strength and statistical power and to better evaluate potential causal links between kidney function and heart failure subtypes in populations of African ancestry.
In this preprint posted July 17, 2026, the MR analysis using African ancestry instruments and Million Veteran Program heart failure subtype data produced no evidence of a causal effect of genetically predicted eGFR on either HFpEF or HFrEF. Because the analysis was constrained by instrument strength and power, the null results are interpreted as inconclusive. The authors call for larger, ancestry-matched GWAS and MR analyses to more definitively assess whether kidney function causally contributes to heart failure subtypes in African ancestry populations.
Competing interests were declared as none, and the authors confirmed adherence to ethical guidelines and participant consent processes in the submission declarations.