ABSTRACT
Background
APOL1 risk variants are strongly associated with kidney disease in individuals of recent African ancestry, but their haemodynamic effects remain incompletely characterised, particularly in African dialysis populations. We evaluated the association between APOL1 variants and blood pressure phenotypes in a South African haemodialysis cohort.
Methods
In this cross-sectional study, 115 adults receiving maintenance haemodialysis in Pretoria were included. All participants were self-identified Black Africans. APOL1 genotyping (G1: rs73885319, rs60910145; G2: rs71785313) was performed, and genetic models were analysed as additive, dominant, and recessive. Pre-dialysis blood pressure was the mean of five consecutive measurements. Outcomes included systolic blood pressure (SBP), diastolic blood pressure (DBP), mean arterial pressure (MAP), and pulse pressure (PP). Associations were assessed using linear mixed-effects models with dialysis site as a random effect, with sequential adjustment for age, sex, and antihypertensive class burden.
Results
APOL1 risk allele distribution was 53.0% zero, 36.5% one, and 10.4% two alleles. There were no consistent associations between APOL1 genotype and SBP, DBP, or MAP in univariate analyses. In multivariable models, SBP was significantly lower in APOL1 risk allele carriers under additive and dominant models. By contrast, PP was consistently lower in APOL1 risk allele carriers and remained significant after full adjustment (additive model, one vs. zero: β −7.772, 95% confidence interval (CI) −13.009 to −2.535, P = 0.004; dominant model: β −7.69 95% CI −12.622 to −2.757, P = 0.003). No consistent associations were observed for DBP or MAP.
Conclusions
In this cohort of Black African haemodialysis patients, APOL1 risk variants were associated with lower PP and, after adjustment, lower SBP, without differences in DBP or MAP. These findings suggest a distinct haemodynamic phenotype characterised by reduced arterial pulsatility and highlight the need for further mechanistic and prospective studies, particularly in the context of emerging APOL1-targeted therapies.
Keywords: African ancestry, APOL1, blood pressure, chronic kidney disease, haemodialysis
Graphical Abstract
Graphical Abstract.
KEY LEARNING POINTS.
What was known:
APOL1 risk variants are strongly associated with kidney disease in people of recent African ancestry, but their haemodynamic effects remain incompletely characterised.
Previous studies have suggested possible associations between APOL1 risk variants, blood pressure traits, cardiovascular phenotypes, and haemodialysis survival, mainly in African American populations.
There are limited data on APOL1 and blood pressure phenotypes in African patients receiving maintenance haemodialysis, despite the African origin of APOL1 risk variants.
This study adds:
In a South African haemodialysis cohort, APOL1 risk allele carriage was associated with consistently lower pulse pressure across additive and dominant genetic models.
APOL1 risk allele carriers also had lower systolic blood pressure after adjustment for age, sex, dialysis site, and antihypertensive class burden.
No consistent associations were observed between APOL1 genotype and diastolic blood pressure or mean arterial pressure, suggesting a phenotype related to pulsatile rather than global blood pressure effects.
Potential impact:
These findings suggest that APOL1 risk variants may influence vascular or pulsatile haemodynamics in dialysis patients, beyond their established role in kidney disease.
APOL1 status may become relevant to cardiovascular phenotyping and risk interpretation in African haemodialysis populations.
Future studies of APOL1 targeted therapies should include cardiovascular and haemodynamic outcomes, particularly in African populations with a high prevalence of APOL1 risk variants.
INTRODUCTION
APOL1 nephropathy is a concept introduced to suggest a common pathway to chronic kidney disease (CKD), mediated by specific variants in the apolipoprotein L1 (APOL1) gene. These variants, termed G1 and G2, are found almost exclusively in people of recent African ancestry, and are thought to explain some of the excess renal disease seen in these populations [1, 2]. APOL1 appears to have pleiotropic effects, with the G1 and G2 alleles protecting against Trypanosoma infection, but increasing risk of many diverse renal diseases; including focal segmental glomerulonephritis, lupus collapsing glomerulopathy, accelerated diabetic and hypertensive renal disease; and decreased kidney transplant graft survival [1, 3]. While initially protective against African sleeping sickness, an additional clinical ‘second hit’ is thought to be necessary to develop renal disease [4]. These pleiotropic effects may extend beyond the kidney to include vascular and endothelial function.
As renal disease is more common in populations of African ancestry, APOL1 nephropathy is often proposed as a key contributor to the increased risk of renal disease seen in these populations [5–7]. The general consensus is that the G1 and G2 alleles are toxic gain of function variants, with two risk variants (G1/G1, G1/G2, or G2/G2) needed to confer the increased renal risk, however, there is some evidence that a single APOL1 risk allele can cause disease [2, 5, 8]. Additionally, another APOL1 variant (p.N264K, rs73885316) has been shown to ‘rescue’ the toxic effect of the G2 allele [9].
While there is strong evidence for APOL1 in many renal diseases, the haemodynamic consequences of APOL1 risk variants, particularly in patients receiving dialysis, remain poorly characterised. In a study of over 9000 self-identified African Americans with relatively preserved estimated glomerular filtration rate (eGFR) (>80 ml/min/1.73 m2), a significant association between increased blood pressure (systolic and diastolic) and APOL1 risk variants was found [6]. Of note, the association was more notable in the younger (20–29 years) group, than in the older groups, and the association was usually attenuated (but did not disappear) with decreasing eGFR [6]. In another study, African American patients with a higher baseline eGFR (>100 ml/min/173 m2) revealed no significant difference between blood pressure and APOL1 risk variants. However, in this study, there was a significantly improved systolic blood pressure in response to candesartan in patients with more than one APOL1 risk alleles [10]. Additionally, in a prospective study, Ku et al. assessed the effect of strict blood pressure control on mortality, stratified by APOL1 risk allele status [11]. In the strict blood pressure control group with two APOL1 risk alleles, there was a significant reduction in the mortality risk (hazard ration 0.58, P = 0.03) [11]. Of note, at baseline, the two APOL1 risk allele group had a significantly lower systolic and diastolic blood pressure compared to those with one or less risk alleles, despite having a significantly lower eGFR.
Despite accelerating patients towards requiring renal replacement therapy, APOL1 risk variants may provide patients with an unexpected benefit. In non-diabetic haemodialysis patients of African American ancestry, a significant survival advantage was seen in patients with two risk alleles, although the exact mechanism of this advantage was not identified [12]. By contrast, Chen et al. found no mortality benefit for APOL1 risk alleles, but did see a significant benefit for lower systolic blood pressure, lower pulse pressure, less left ventricular hypertrophy, and lower coronary artery calcification in patients with two risk alleles [13]. While this study did not find any mortality benefit despite having clinical benefits, the follow up on this study was relatively short (mean 2.5 years) and may not have been powered enough to detect a mortality benefit [13].
APOL1 risk alleles have been consistently associated with multiple renal diseases, proteinuria and accelerated eGFR decline, in a recessive pattern (two risk alleles) in populations worldwide that are of recent African ancestry [1]. Additionally, there is a less well-defined signal that APOL1 risk alleles may be associated with improved blood pressure (pre and dialysis requiring patients) as well as a possible haemodialysis survival benefit, but in an additive manner (one or two risk alleles) [10–12]. However, despite the African origin of APOL1 risk variants, these studies have been conducted predominantly in African American populations, with limited data from African cohorts. This study aimed to evaluate the association between APOL1 risk variants and haemodynamic parameters, including systolic blood pressure and pulse pressure, in a cohort of self-identified Black African patients receiving maintenance haemodialysis, and to assess interactions with antihypertensive treatment.
MATERIALS AND METHODS
Study design and population
This cross-sectional analysis was conducted within a cohort of adults with end-stage kidney disease receiving maintenance haemodialysis. Patients were recruited from multiple dialysis units in the greater Pretoria region. Patients were recruited using a non-random, centre-based consecutive sampling approach. At each participating dialysis unit, all eligible prevalent haemodialysis patients were approached for participation during the recruitment period. Patients were included if they met the inclusion criteria and provided written informed consent. Eligible participants were aged ≥18 years, had been on haemodialysis for at least three months, and were all self-identified Black Africans. Written informed consent was obtained from all participants, and ethical approval was granted by the University of Pretoria Research Ethics Committee (147/2023 and 147/2020).
Data collection and blood pressure assessment
Demographic and clinical data, including age, sex, dialysis site, and antihypertensive medication use, were obtained from routine clinical records. Pre-dialysis blood pressure was defined as the mean of five consecutive measurements per participant for Systolic Blood Pressure (SBP) and Diastolic Blood Pressure (DBP). Derived haemodynamic indices included mean arterial pressure (MAP), calculated as DBP plus one-third of the pulse pressure (PP), and PP, calculated as the difference between SBP and DBP.
Genotyping and genetic models
Peripheral blood samples were collected in EDTA tubes, and genomic DNA was extracted using sodium perchlorate and chloroform instead of phenol [14]. APOL1 genotyping was performed using Sanger sequencing of exon 7 (BigDyeTM Terminator v3.1 Cycle Sequencing Kit, Applied BiosystemsTM, Thermo Fischer Scientific, USA) using inhouse designed primers (Forward - CAG CTG AAA GCG GTG AAC AG and Reverse - GGC ATA TCT CTC CTG GTG GC), or whole-exome sequencing (Origenes, Western Cape, South Africa). Variants analysed included G1 (rs73885319 A > G and rs60910145 T > G) and G2 (rs71785313 deletion of the six-base pair sequence TTATAA).
APOL1 risk status was defined using three inheritance models: additive (0, 1, or 2 risk alleles), dominant (0 vs. 1 and 2 alleles), and recessive (0 and 1 vs. 2 alleles).
Antihypertensive therapy
Antihypertensive medications were classified by pharmacological class. Antihypertensive drug burden was defined as the total number of drug classes prescribed and included as a grouped covariate in regression analyses.
Statistical analysis
Continuous variables were presented as mean (standard deviation) or median (interquartile range), as appropriate, and categorical variables as number (percentage). Associations between APOL1 variants and SBP, DBP, MAP, and PP were assessed using linear mixed-effects models. Dialysis site was included as a random intercept to account for clustering. Sequential modelling was performed for each outcome: an unadjusted model including APOL1 genetic model alone, a partially adjusted model including age and sex, and a fully adjusted model including age, sex, and antihypertensive class burden. Outcomes were analysed separately due to collinearity between blood pressure indices. Model estimates are presented as regression coefficients with 95% confidence intervals. All statistical tests were two-sided, with a P-value < 0.05 considered statistically significant. Analyses were performed using RStudio (version 2025.05.1) [15].
RESULTS
Cohort characteristics
A total of 115 patients with dialysis-dependent CKD were included. The cohort was predominantly male (67.0%), with a mean age of 45.9 ± 13.6 years. Blood pressure was elevated across all measures, with a high antihypertensive burden (Table 1).
Table 1:
Cohort characteristics.
| Variable | Total | Female | Male | p |
|---|---|---|---|---|
| Number, n (%) | 115 (100.0%) | 38 (33.0%) | 77 (67.0%) | – |
| Age, mean (SD) | 45.9 (13.6) | 44.2 (14.6) | 46.7 (13.2) | 0.380 |
| Private sector, n (%) | 56 (48.7%) | 16 (42.1%) | 40 (52.0%) | 0.321 |
| Public sector, n (%) | 59 (51.3%) | 22 (57.9%) | 37 (48.1%) | |
| SBP, mean (SD) | 148.6 (21.0) | 147.0 (21.1) | 149.3 (21.1) | 0.603 |
| DBP, mean (SD) | 87.4 (14.8) | 85.9 (15.9) | 88.2 (14.3) | 0.487 |
| MAP, mean (SD) | 107.8 (15.8) | 106.3 (16.3) | 108.5 (15.6) | 0.507 |
| PP, mean (SD) | 61.2 (14.1) | 61.1 (15.5) | 61.2 (13.5) | 0.988 |
| Antihypertensive medication classes | ||||
| Classes per patient, median (IQR) | 3.0 (2.0–5.0) | 3.0 (2.0–5.0) | 3.0 (1.0–5.0) | 0.545 |
| 0–1 classes, n (%) | 28 (24.4%) | 8 (21.1%) | 20 (26.0%) | 0.831 |
| 2–3 classes, n (%) | 36 (31.3%) | 12 (31.6%) | 24 (31.2%) | |
| ≥4 classes, n (%) | 51 (44.4%) | 18 (47.4%) | 33 (42.9%) | |
| APOL1 additive model | ||||
| Zero risk alleles | 61 (53.0%) | 23 (60.5%) | 38 (49.4%) | 0.340 |
| One risk allele | 42 (36.5%) | 13 (34.2%) | 29 (37.7%) | |
| Two risk alleles | 12 (10.4%) | 2 (5.3%) | 10 (13.0%) | |
| APOL1 dominant model | ||||
| Zero risk alleles | 61 (53.0%) | 23 (60.5%) | 38 (49.4%) | 0.259 |
| One and two risk alleles | 54 (47.0%) | 15 (39.5%) | 39 (50.7%) | |
| APOL1 recessive model | ||||
| Zero and one risk alleles | 103 (89.6%) | 36 (94.7%) | 67 (87.0%) | 0.332 |
| Two risk alleles | 12 (10.4%) | 2 (5.3%) | 10 (13.0%) |
SD: standard deviation; IQR: interquartile range; n: number. Blood pressure is reported in mmHg. MAP was calculated as DBP + 1/3(SBP − DBP), and PP as SBP − DBP. Continuous variables are presented as mean (SD) or median (IQR), and categorical variables as number (percentage). P-values reflect comparisons between female and male participants using appropriate tests (t-test, Wilcoxon rank-sum, Chi-square, or Fisher’s exact). Antihypertensive burden is summarised by number of classes and grouped as 0–1, 2–3, and ≥4 classes. APOL1 models are defined as additive (0, 1, 2 risk alleles), dominant (0 vs. ≥1), and recessive (≤1 vs. 2).
APOL1 distribution
APOL1 genotype distribution was as follows: in the additive model, 53.0% of participants had zero risk alleles, 36.5% had one risk allele, and 10.4% had two risk alleles. Under the dominant model, 53.0% had no risk alleles and 47.0% carried one or more risk alleles. In the recessive model, 89.6% had zero or one risk allele and 10.4% had two risk alleles. Genotype frequencies for the APOL1 risk variants rs73885319 and rs60910145 (G1) were similar, with 81.7% homozygous non-risk variant, 17.4% heterozygous, and 0.9% homozygous risk variant for both loci, and were in Hardy–Weinberg equilibrium (P > 0.999 for both). For the G2 (rs71785313), genotype, frequencies were 68.7% homozygous non-risk variant, 24.3% heterozygous, and 7.0% homozygous risk variant, with significant deviation from Hardy–Weinberg equilibrium observed (P = 0.031). Linkage disequilibrium analysis demonstrated near-complete linkage between rs73885319 and rs60910145 (D′ = 0.999, r² = 0.998), consistent with their joint definition of the G1 haplotype. By contrast, rs71785313 (G2) showed low correlation with both rs73885319 and rs60910145 (r² = 0.025 for both), despite high D′ (0.996), indicating low allele frequency-driven linkage. There were no significant differences in age, sex, or antihypertensive treatment burden across APOL1 groups in any genetic model.
Between-group comparisons
Blood pressure parameters were compared across APOL1 genetic models using Kruskal–Wallis testing. There were no significant differences in SBP, DBP, or MAP across any genetic model. By contrast, PP differed significantly across APOL1 groups, with consistently lower values observed in carriers of APOL1 risk alleles. This effect was most pronounced in the dominant model (Table 2).
Table 2:
Blood pressure across APOL1 genetic models.
| Outcome | Model | Group 1 | Group 2 | Group 3 | p |
|---|---|---|---|---|---|
| SBP (mmHg) | Additive | 151.8 (18.8) | 145.0 (25.2) | 143.4 (13.0) | 0.195 |
| Dominant | 151.8 (18.8) | 144.7 (23.1) | – | 0.072 | |
| Recessive | 149.1 (21.7) | 143.4 (13.0) | – | 0.410 | |
| DBP (mmHg) | Additive | 87.4 (14.2) | 88.0 (17.0) | 85.5 (8.9) | 0.955 |
| Dominant | 87.4 (14.2) | 87.5 (15.6) | – | 0.915 | |
| Recessive | 87.6 (15.3) | 85.5 (8.9) | – | 0.760 | |
| MAP (mmHg) | Additive | 109.0 (14.3) | 107.0 (19.2) | 104.8 (9.0) | 0.798 |
| Dominant | 109.0 (14.3) | 106.5 (17.5) | – | 0.517 | |
| Recessive | 108.1 (16.3) | 104.8 (9.0) | – | 0.692 | |
| PP (mmHg) | Additive | 64.4 (14.6) | 57.0 (12.7) | 57.9 (11.3) | 0.026 |
| Dominant | 64.4 (14.6) | 57.2 (12.3) | – | 0.007 | |
| Recessive | 61.5 (14.3) | 57.9 (11.3) | – | 0.486 |
Additive: Group 1 = 0 risk alleles, Group 2 = 1 risk allele, Group 3 = 2 risk alleles. Dominant: Group 1 = 0 risk alleles, Group 2 = 1 or 2 risk alleles. Recessive: Group 1 = 0 or 1 risk alleles, Group 2 = 2 risk alleles. Bold: P < 0.05.
Blood pressure and APOL1 genetic models
Univariate analysis
Increasing age was associated with lower DBP and MAP and higher PP, while sex was not associated with any blood pressure parameter (Supplementary data). There were no significant univariate associations between APOL1 genotype and SBP, DBP, or MAP across additive, dominant, or recessive models. By contrast, PP differed significantly by APOL1 status, with lower PP observed in carriers under additive (one vs. zero) and dominant models (Table 3).
Table 3:
Univariate associations between APOL1 models and blood pressure.
| Outcome | Model | Comparison | Beta (95% CI) | p |
|---|---|---|---|---|
| SBP | Additive | One vs. zero | −6.78 (−15.40, 1.84) | 0.122 |
| Two vs. zero | −8.38 (−22.56, 5.80) | 0.244 | ||
| Dominant | One/two vs. zero | −7.11 (−15.13, 0.91) | 0.082 | |
| Recessive | Two vs. zero/one | −5.72 (−19.58, 8.14) | 0.415 | |
| DBP | Additive | One vs. zero | 0.55 (−5.60, 6.70) | 0.860 |
| Two vs. zero | −1.92 (−12.03, 8.18) | 0.707 | ||
| Dominant | One/two vs. zero | 0.04 (−5.69, 5.76) | 0.990 | |
| Recessive | Two vs. zero/one | −2.14 (−11.91, 7.63) | 0.665 | |
| MAP | Additive | One vs. zero | −1.89 (−8.44, 4.65) | 0.568 |
| Two vs. zero | −4.08 (−14.84, 6.69) | 0.454 | ||
| Dominant | One/two vs. zero | −2.35 (−8.44, 3.75) | 0.447 | |
| Recessive | Two vs. zero/one | −3.33 (−13.75, 7.07) | 0.527 | |
| PP | Additive | One vs. zero | −7.33 (−12.99, −1.67) | 0.012 |
| Two vs. zero | −6.46 (−15.76, 2.85) | 0.172 | ||
| Dominant | One/two vs. zero | −7.15 (−12.41, −1.89) | 0.008 | |
| Recessive | Two vs. zero/one | −3.59 (−12.85, 5.69) | 0.445 |
mmHg: millimetres of mercury; Beta: regression coefficient representing the mean difference in outcome between groups; CI: confidence interval; p: P value indicating statistical significance. Model definitions: Additive model compares 0, 1, and 2 APOL1 risk alleles; Dominant model compares 0 versus 1 or 2 risk alleles; Recessive model compares 0 or 1 versus 2 risk alleles. Negative beta values indicate lower blood pressure relative to the reference group. Bold: P < 0.05.
Multivariable analysis
 |
There were no consistent associations between APOL1 genotype and DBP or MAP across models. SBP was lower in APOL1 risk allele carriers under additive and dominant models. Similarly, PP was lower in APOL1 risk allele carriers, again under additive and dominant models, but not the recessive model (Table 4).
Table 4:
Multivariable associations (base vs. fully adjusted models).
| Base model | Fully adjusted model | |||||
|---|---|---|---|---|---|---|
| Outcome | Model | Comparison | Beta (95% CI) | p | Beta (95% CI) | p |
| SBP | Additive | One vs. zero | −10.06 (−18.20, −1.93) | 0.016 | −8.41 (−16.27, −0.54) | 0.036 |
| Two vs. zero | −11.76 (−24.93, 1.41) | 0.079 | −11.81 (−24.81, 1.18) | 0.074 | ||
| Dominant | One/two vs. zero | −10.43 (−18.04, −2.83) | 0.008 | −9.11 (−16.52, −1.70) | 0.017 | |
| Recessive | Two vs. zero/one | −7.62 (−20.78, 5.56) | 0.254 | −8.12 (−20.98, 4.75) | 0.214 | |
| DBP | Additive | One vs. zero | −1.94(−7.02, 3.15) | 0.452 | −0.99 (−5.98, 3.99) | 0.694 |
| Two vs. zero | −4.60 (−12.78, 3.57) | 0.267 | −4.54 (−12.71, 3.63) | 0.273 | ||
| Dominant | One/two vs. zero | −2.53 (−7.27, 2.22) | 0.293 | −1.75 (−6.44, 2.94) | 0.462 | |
| Recessive | Two vs. zero/one | −3.84 (−11.80, 4.12) | 0.341 | −4.13 (−12.04, 3.78) | 0.303 | |
| MAP | Additive | One vs. zero | −4.66 (−10.42, 1.09) | 0.111 | −3.45 (−9.03, 2.13) | 0.223 |
| Two vs. zero | −7.01 (−16.28, 2.26) | 0.137 | −6.98 (−16.16, 2.20) | 0.134 | ||
| Dominant | One/two vs. zero | −5.18 (−10.55, 0.19) | 0.058 | −4.20 (−9.45, 1.06) | 0.117 | |
| Recessive | Two vs. zero/one | −5.14 (−14.27, 3.99) | 0.266 | −5.52 (−14.48, 3.44) | 0.224 | |
| PP | Additive | One vs. zero | −8.28 (−13.58, −2.98) | 0.002 | −7.77 (−13.01, −2.54) | 0.004 |
| Two vs. zero | −7.13(−15.76, 1.50) | 0.104 | −7.36 (−16.02, 1.31) | 0.095 | ||
| Dominant | One/two vs. zero | −8.04 (−13.001, −3.08) | 0.002 | −7.70 (−12.62, −2.76) | 0.003 | |
| Recessive | Two vs. zero/one | −3.60 (−12.36, 5.17) | 0.418 | −3.90 (−12.64, 4.84) | 0.378 | |
Beta: adjusted mean difference versus reference; p: P value. Models: Additive (0 vs. 1 vs. 2 alleles, reference = 0); Dominant (0 vs. 1 or 2 alleles, reference = 0). Comparisons: One versus zero: 1 vs. 0 alleles; Two versus zero: 2 versus 0 alleles; One/two versus zero: ≥1 versus 0 alleles. Negative beta indicates lower blood pressure relative to the reference after adjustment. Bold: P < 0.05.
 |
Adjustment for antihypertensive class burden did not materially alter the association between APOL1 and SBP (Table 4). Lower SBP remained evident under additive and dominant models, reaching statistical significance for individuals with one risk allele and for carriers in the dominant model, while the comparison for two versus zero alleles remained non-significant. For PP, the association with APOL1 was consistent across models. Individuals with one risk allele in the additive model and carriers in the dominant model demonstrated significantly lower PP in both base and fully adjusted analyses.
Summary of findings
APOL1 genotype was not associated with DBP or MAP across genetic models. SBP was lower in APOL1 risk allele carriers under additive and dominant models, with statistically significant differences observed in fully adjusted analyses. PP was consistently lower in APOL1 risk allele carriers, particularly under additive and dominant models, and this association was present in both base and fully adjusted models.
DISCUSSION
In this cross-sectional cohort of Black African haemodialysis patients, APOL1 risk variants were associated with a consistent haemodynamic signal characterised by lower PP. It was significantly reduced in APOL1 risk allele carriers under additive and dominant models, with this effect observed in Kruskal–Wallis, univariate, and multivariate regression analyses; driven primarily by individuals with one risk allele. By contrast, SBP was lower in APOL1 risk allele carriers only in multivariable models, with significant reductions seen for one versus zero alleles in the additive model and in the dominant model. No consistent associations were observed for MAP or DBP, across genetic models. To our knowledge, this represents the first analysis of APOL1 and blood pressure phenotypes in an African dialysis population [16]. Collectively, these findings support a model in which APOL1 risk variants are associated with a distinct haemodynamic phenotype, predominantly reflected by lower PP rather than uniform changes across conventional blood pressure measures.
SBP demonstrated a context-dependent association, with no significant differences observed in univariate analyses but significant reductions emerging after multivariable adjustment under additive and dominant models. This pattern suggests that the relationship between APOL1 and SBP may be partially masked by confounding or treatment-related effects. Prior studies have reported differential blood pressure responses in APOL1 risk allele carriers, including greater systolic blood pressure reduction in response to angiotensin receptor blockade, supporting a potential interaction between APOL1 status and antihypertensive therapy [10]. This is potentially a significant confounder as 63.5% of our study populations were taking angiotensin receptor blockers or angiotensin converting enzyme inhibitors. Additionally, lower baseline blood pressure has been observed in individuals with two APOL1 risk alleles despite more advanced kidney disease, further supporting a potential intrinsic haemodynamic effect [11]. Together, these findings suggest that APOL1-associated differences in SBP may become apparent only after accounting for clinical and treatment-related heterogeneity.
The most consistent and biologically coherent finding in this study was the reduction in PP among APOL1 risk allele carriers. PP is a well-established surrogate marker of arterial stiffness and pulsatile haemodynamic load, and is strongly associated with cardiovascular risk in both the general and dialysis populations [17, 18]. The observed reduction in PP in APOL1 carriers, which remained robust across additive and dominant models and after adjustment for antihypertensive treatment burden, suggests a potential difference in vascular compliance or arterial properties. This is consistent with prior observations in haemodialysis cohorts demonstrating lower PP, reduced left ventricular hypertrophy, and lower coronary artery calcification in APOL1 high-risk individuals, despite their increased susceptibility to kidney disease [13]. These findings support the presence of a distinct haemodynamic phenotype in APOL1 risk allele carriers, characterised by reduced arterial pulsatility, and may provide a potential mechanistic link between APOL1 genotype and potentially favourable cardiovascular risk profiles in this population. A possible explanation is that the adverse APOL1-associated renal and hypertension-related phenotype may predominate before kidney failure, masking any favourable haemodynamic effect. In dialysis-dependent CKD, this renal phenotype may no longer be the dominant clinical signal, allowing a lower-pulse-pressure phenotype to become apparent.
The combined findings for SBP and PP suggest that APOL1 risk variants may influence vascular haemodynamics beyond alterations in blood pressure. While SBP and PP were lower in APOL1 risk allele carriers, DBP and MAP were not different across genetic models, indicating that the observed effects are unlikely to reflect a global reduction in blood pressure. Rather, this pattern is more consistent with alterations in pulsatile haemodynamics, potentially reflecting differences in arterial stiffness, wave reflection, or vascular compliance [17]. The dissociation between systolic and pulsatile measures on the one hand, and mean or diastolic pressures on the other, supports the concept that APOL1-associated effects may be mediated through vascular properties rather than systemic pressure regulation alone. This is particularly relevant in the dialysis population, where arterial stiffness and altered vascular compliance are key determinants of cardiovascular risk.
The observed haemodynamic pattern is consistent with APOL1 as a pleiotropic gene shaped by evolutionary selection. APOL1 risk variants (G1 and G2) arose under positive selection due to protection against Trypanosoma brucei infection, but confer increased risk of kidney disease, including focal segmental glomerulosclerosis and hypertensive nephropathy [3, 19]. Beyond the kidney, emerging evidence suggests broader effects on vascular and endothelial function. In pre-eclampsia, APOL1 genotype has been linked to placental and vascular pathology, with associations reported for fetal genotype, maternal variants in African populations, and maternal–fetal interactions, although large cohort data demonstrate heterogeneity and, in some cases, no association [20–23]. In addition, APOL1 risk variants have been associated with differences in blood pressure traits and cardiovascular phenotypes in non-dialysis populations [6]. In this context, the consistent reduction in pulse pressure and the adjusted association with lower SBP observed in this study may reflect underlying differences in vascular compliance or arterial function in APOL1 risk allele carriers, rather than a simple effect on systemic blood pressure regulation.
The relationship between APOL1 and haemodynamic measures should be interpreted in the context of potential confounding, particularly confounding by indication. Antihypertensive therapy was common in this cohort, with 73 patients (63.5%) receiving ACE inhibitors or angiotensin receptor blockers. Greater treatment burden, reflected by increasing numbers of antihypertensive classes, was strongly associated with higher SBP, DBP, and MAP, supporting the presence of confounding by indication, whereby patients with more severe hypertension receive more intensive therapy (supplemental data). SBP differences were not evident in unadjusted comparisons but became apparent after multivariable adjustment, first in the base model and persisting after adjustment for antihypertensive class burden. This suggests that the association between APOL1 and SBP may be partially obscured by treatment and clinical heterogeneity rather than fully explained by treatment burden alone. However, adjustment for antihypertensive class burden cannot fully account for differences in treatment intensity, adherence, or drug specific effects. Prior studies have suggested that APOL1 risk allele carriers may exhibit differential responses to renin–angiotensin system blockade, including greater reductions in systolic blood pressure [18]. This raises the possibility that pharmacological effects may interact with APOL1 status, contributing to the unmasking of SBP differences after adjustment. Residual confounding from unmeasured factors, including volume status, dialysis adequacy, and vascular calcification, is also likely and may influence both blood pressure and treatment requirements in this population.
These findings may have potential implications for emerging APOL1-targeted therapies. Pharmacological inhibition of APOL1, such as with Inaxaplin, has demonstrated reductions in proteinuria in individuals with APOL1-mediated kidney disease [24]. However, if APOL1 risk variants are associated with a haemodynamic pattern characterised by lower PP and reduced SBP under certain conditions, inhibition of APOL1 function may alter these vascular effects. While speculative, this raises the possibility of a trade-off between renal benefit and changes in vascular or haemodynamic profiles. These considerations highlight the importance of comprehensive cardiovascular phenotyping in future studies of APOL1-targeted therapies, particularly in African populations where both the prevalence of APOL1 risk variants and the burden of cardiovascular disease are high.
This study has several limitations. The cross-sectional design precludes causal inference and limits interpretation to associations. The sample size was modest, particularly for the recessive APOL1 model, which likely had reduced power to detect smaller effects. Residual confounding is likely, despite adjustment for key variables, particularly given the complexity of antihypertensive treatment patterns and the potential for confounding by indication. Important physiological factors, including volume status, dialysis adequacy, residual kidney function, vascular calcification, and direct measures of arterial stiffness, were not available and may have influenced the observed haemodynamic relationships. Although participants were recruited from 11 dialysis centres, all centres were located within the greater Pretoria region, which may limit generalisability to other African dialysis populations.
This study has several strengths. It represents one of the first evaluations of APOL1-associated haemodynamic phenotypes in a self-identified Black African haemodialysis population. The inclusion of patients from multiple dialysis centres enhances the representativeness of the cohort within the regional context. Blood pressure measurements were based on repeated pre-dialysis readings, improving reliability compared to single measurements. In addition, the use of multiple APOL1 genetic models and mixed-effects modelling allowed for a comprehensive assessment of genotype–phenotype associations while accounting for clustering by dialysis site.
In this cohort of Black African haemodialysis patients, APOL1 risk variants were associated with a distinct haemodynamic phenotype characterised by consistently lower PP and, after adjustment, lower SBP, without the corresponding differences in DBP or MAP. These findings suggest that APOL1 may influence vascular or pulsatile haemodynamics beyond its established role in kidney disease. Further prospective and mechanistic studies are needed to clarify the underlying pathways and to determine the clinical implications of these findings, particularly in the context of emerging APOL1-targeted therapies.
Supplementary Material
ACKNOWLEDGEMENTS
The authors acknowledge National Renal Care for their assistance with data extraction for a subset of patients included in this study. No financial support, remuneration, or other financial transactions were involved in this assistance.
Contributor Information
Wesley van Hougenhouck-Tulleken, University of Pretoria, Departments of Internal Medicine and Genetics, Pretoria, South Africa; Sefako Makgatho Health Sciences University, Department of Internal Medicine, Ga-Rankuwa, South Africa.
Paul Rheeder, University of Pretoria, Departments of Internal Medicine and Genetics, Pretoria, South Africa.
Elizabeth J van Rensburg, Departments of Human Genetics, University of Pretoria, Pretoria, South Africa.
CONFLICT OF INTEREST STATEMENT
None declared.
FUNDING
None declared.
DATA AVAILABILITY STATEMENT
The data underlying this article are available in the article.
REFERENCES
- 1. Friedman DJ, Pollak MR. APOL1 nephropathy: from genetics to clinical applications. Clin J Am Soc Nephrol. 2021;16:294–303. 10.2215/CJN.15161219 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Ojo AO, Adu D, Bramham K et al. APOL1 kidney disease: conclusions from a Kidney Disease: Improving Global Outcomes (KDIGO) Controversies Conference. Kidney Int. 2025;108:763–79. 10.1016/j.kint.2025.05.017 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Genovese G, Friedman DJ, Ross MD et al. Association of trypanolytic APOL1 variants with kidney disease in African Americans. Science. 2010;329:841–5. 10.1126/science.1193032 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Daneshpajouhnejad P, Kopp JB, Winkler CA et al. The evolving story of apolipoprotein L1 nephropathy: the end of the beginning. Nat Rev Nephrol. 2022;18:307–20. 10.1038/s41581-022-00538-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Pollak MR, Friedman DJ. APOL1 and APOL1-associated kidney disease: a common disease, an unusual disease gene—proceedings of the henry shavelle professorship. Glomerular Dis. 2023;3:75–87. 10.1159/000529227 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Nadkarni GN, Galarneau G, Ellis SB et al. Apolipoprotein L1 variants and blood pressure traits in African Americans. J Am Coll Cardiol. 2017;69:1564–74. 10.1016/j.jacc.2017.01.040 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Ekulu PM, Nkoy AB, Betukumesu DK et al. APOL1 risk genotypes are associated with early kidney damage in children in sub-Saharan Africa. Kidney Int Rep. 2019;4:930–8. 10.1016/j.ekir.2019.04.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Itoku A, Isaac J, Wilson S et al. APOL1 nephropathy risk variants through the life course: a review. Am J Kidney Dis. 2024;84:102–110. 10.1053/j.ajkd.2023.12.014 [DOI] [PubMed] [Google Scholar]
- 9. Gupta Y, Friedman DJ, McNulty MT et al. Strong protective effect of the APOL1 p.N264K variant against G2-associated focal segmental glomerulosclerosis and kidney disease. Nat Commun. 2023;14:7836. 10.1038/s41467-023-43020-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Cunningham PN, Wang Z, Grove ML et al. Hypertensive APOL1 risk allele carriers demonstrate greater blood pressure reduction with angiotensin receptor blockade compared to low risk carriers. PLoS ONE. 2019;14:e0221957. 10.1371/journal.pone.0221957 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Ku E, Lipkowitz MS, Appel LJ et al. Strict blood pressure control associates with decreased mortality risk by APOL1 genotype. Kidney Int. 2017;91:443–50. 10.1016/j.kint.2016.09.033 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Ma L, Langefeld CD, Comeau ME et al. APOL1 renal-risk genotypes associate with longer hemodialysis survival in prevalent nondiabetic African American patients with end-stage renal disease. Kidney Int. 2016;90:389–95. 10.1016/j.kint.2016.02.032 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Chen TK, Fitzpatrick J, Winkler CA et al. APOL1 risk variants and subclinical cardiovascular disease in incident hemodialysis patients. Kidney Int Rep. 2021;6:333–41. 10.1016/j.ekir.2020.11.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Johns MB Jr, Paulus-Thomas JE. Purification of human genomic DNA from whole blood using sodium perchlorate in place of phenol. Anal Biochem. 1989;180:276–8. 10.1016/0003-2697(89)90430-2 [DOI] [PubMed] [Google Scholar]
- 15. R Core Team . R: A Language and Environment for Statistical Computing. Vienna: R Foundation for Statistical Computing, 2024. https://www.R-project.org/ [Google Scholar]
- 16. Gbadegesin RA, Ulasi I, Ajayi S et al. APOL1 bi- and monoallelic variants and chronic kidney disease in West Africans. N Engl J Med. 2025;392:228–38. 10.1056/NEJMoa2404211 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Blacher J, Guerin AP, Pannier B et al. Impact of aortic stiffness on survival in end-stage renal disease. Circulation. 1999;99:2434–9. 10.1161/01.CIR.99.18.2434 [DOI] [PubMed] [Google Scholar]
- 18. Lipkowitz MS, Freedman BI, Langefeld CD et al. Apolipoprotein L1 gene variants associate with hypertension-attributed nephropathy and the rate of kidney function decline in African Americans. Kidney Int. 2013;83:114–20. 10.1038/ki.2012.263 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Rosset S, Tzur S, Behar DM et al. The population genetics of chronic kidney disease: insights from the MYH9–APOL1 locus. Nat Rev Nephrol. 2011;7:313–26. 10.1038/nrneph.2011.52 [DOI] [PubMed] [Google Scholar]
- 20. Reidy KJ, Hjorten RC, Simpson CL et al. Fetal-not maternal-APOL1 genotype associated with risk for pre-eclampsia in those with African Ancestry. Am Hum Genet. 2018;103:367–76. 10.1016/j.ajhg.2018.08.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Thakoordeen-Reddy S, Winkler C, Moodley J et al. Maternal variants within the apolipoprotein L1 gene are associated with pre-eclampsia in a South African cohort of African ancestry. Eur J Obstet Gynecol Reprod Biol. 2020;246:129–33. 10.1016/j.ejogrb.2020.01.034 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Hong X, Rosenberg AZ, Zhang B et al. Joint associations of maternal-fetal APOL1 genotypes and maternal country of origin with pre-eclampsia risk. Am J Kidney Dis. 2021;77:879–888.e1. 10.1053/j.ajkd.2020.10.020 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Sheehy S, Friedman D, Liu C et al. Association between Apolipoprotein L1 genetic variants and risk of pre-eclampsia and preterm birth among U.S. Black women. Eur J Obstet Gynecol Reprod Biol X. 2025;25:100365. 10.1016/j.eurox.2025.100365. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Egbuna O, Zimmerman B, Manos G et al. Inaxaplin for proteinuric kidney disease in persons with two APOL1 variants. N Engl J Med. 2023;388:969–79. 10.1056/NEJMoa2202396 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data underlying this article are available in the article.