Michelle Redelinghuys

Relationship Between Urinary Salt Excretion and Pulse Pressure and Central Aortic Hemodynamics Independent of Steady State Pressure in the General Population

Although central pulse pressure (PPc) is strongly related to central mean arterial pressure (MAPc), PPc predicts cardiovascular outcomes beyond MAPc. Whether modifiable risk factors for hypertension contribute to PPc and its determinants, independent of MAPc, is uncertain. In 635 randomly recruited participants, we assessed the independent relationship between 24-hour urinary sodium (Na+) or potassium (K+) excretion and brachial artery PP (in office or 24-hour; n=487), PPc, the forward (P1) and augmented (Paug) pressure wave components of PPc, central augmentation index, and determinants of central pressure waves, including aortic pulse wave velocity, effective reflecting distance, and reflective wave transit time. Central dynamics were determined using applanation tonometry of the carotid, femoral, and radial arteries. With adjustments for potential confounders, urinary Na+/K+ was independently associated with in-office, central, and 24-hour PP, as well as Paug, P1, and central augmentation index (P<0.05 to P<0.005). With further adjustments for MAPc (or diastolic BP), urinary Na+/K+ was independently associated with PPc, 24-hour PP, Paug, P1, and central augmentation index (P<0.05 to P=0.005) but not with in-office PP, pulse wave velocity, effective reflecting distance, or reflective wave transit time. In conclusion, in a population of African ancestry, urinary salt excretion is independently related to central and 24-hour PP independent of MAPc or diastolic BP, effects that are attributed to increases in both P1 and Paug but not to pulse wave velocity. Hence, modifying salt intake could influence cardiovascular risk through effects on 24-hour and central PPs, as well as P1 and Paug, independent of steady-state pressure (MAP or diastolic BP) or pulse wave velocity.

Brachial blood pressure-independent relations between radial late systolic shoulder-derived aortic pressures and target organ changes.

Central aortic blood pressure (BP; BPc) predicts outcomes beyond brachial BP. In this regard, the application of a generalized transfer function (GTF) to radial pulse waves for the derivation of BPc is an easy and reproducible measurement technique. However, the use of the GTF may not be appropriate in all circumstances. Although the peak of the second shoulder of the radial waveform (P2) is closely associated with BPc, and, hence, BPc may be assessed without the need for a GTF, whether P2-derived BPc is associated with adverse cardiovascular changes independent of brachial BP is uncertain. Thus, P2- and GTF-derived aortic BPs were assessed using applanation tonometry and SphygmoCor software. Left ventricular mass was indexed for height1.7 (n=678) and carotid intima-media thickness (IMT; n=462) was determined using echocardiography and vascular ultrasound. With adjustments for nurse-derived brachial pulse pressure (PP), P2-derived central PP was independently associated with left ventricular mass indexed for height1.7 (partial r=0.18; P<0.0001) and IMT (partial r=0.40; P<0.0001). These relations were similar to nurse-derived brachial PP-independent relations between GTF-derived central PP and target organ changes (left ventricular mass indexed for height1.7: partial r=0.17, P<0.0001; IMT: partial r=0.37, P<0.0001). In contrast, with adjustments for central PP, nurse-derived brachial PP-target organ relations were eliminated (partial r=−0.21 to 0.05). Twenty-four–hour, day, and night PP-target organ relations did not survive adjustments for nurse-derived brachial BP. In conclusion, central PP derived from P2, which does not require a GTF, is associated with cardiovascular target organ changes independent of brachial BP. Thus, when assessing adverse cardiovascular effects of aortic BP independent of brachial BP, P2-derived measures may complement GTF-derived measures of aortic BP.

Gender-specific contribution of aortic augmentation index to variations in left ventricular mass index in a community sample of African ancestry

Although indices of aortic augmentation derived from radial applanation tonometry are independently associated with adverse cardiovascular effects, whether these relationships are influenced by gender is uncertain. We compared the brachial blood pressure-independent contribution of augmentation index (AIx) to variations in left ventricular mass index (LVMI) in a community sample of 808 participants, 283 of whom were men. Aortic haemodynamics were determined using radial applanation tonometry and SphygmoCor software and LVMI from echocardiography. In men, both AIx derived from aortic augmentation pressure/central aortic pulse pressure (AP/PPc; partial r=0.17, β-coefficient±s.e.m.=0.55±0.20, P<0.01) and AIx derived from the second peak/first peak (P2/P1) of the aortic pulse wave (partial r=0.21, β-coefficient±s.e.m.=0.42±0.12, P<0.0005) were associated with LVM indexed to body surface area (LVMI–BSA). In contrast, in women, neither AIx derived from AP/PPc (partial r=−0.08, β-coefficient±s.e.m.=−0.20±0.11, P=0.08) nor AIx derived from P2/P1 (partial r=−0.06, β-coefficient±s.e.m.=−0.07±0.05, P=0.17) were associated with LVMI–BSA. Both the strength of the correlations (P<0.001 and P<0.0005 with z-statistics) and the slope of the AIx–LVMI relationships (P=0.001 and P<0.0005) were greater in men as compared with women. The lack of relationship between AIx and LVMI was noted in both premenopausal (n=285; AP/PPc vs. LVMI–BSA, partial r=0.01, P=0.95, P2/P1 vs. LVMI–BSA, partial r=0.02, P=0.77), and postmenopausal (n=240; AP/PPc vs. LVMI–BSA, partial r=−0.06, P=0.37, P2/P1 vs. LVMI–BSA, partial r=−0.03, P=0.64) women. Similar differences were noted in the relationships between AIx and LVM indexed to height2.7 in men and women. In conclusion, radial applanation tonometry-derived AIx may account for less of the variation in end-organ changes in women as compared with men.

Lack of Independent Association Between C-Reactive Protein and Central Aortic Hemodynamics in Black Africans with High Risk of Cardiovascular Disease

BACKGROUND As the independent relationship between C-reactive protein (CRP) concentrations and central aortic blood pressures (BP) in populations with high CRP concentrations is uncertain, we aimed to evaluate this question in a community sample, 57% with CRP concentrations >3 mg/l (high cardiovascular risk). METHODS Central aortic hemodynamics (applanation tonometry, SphygmoCor software) and serum ultrasensitive-CRP concentrations were assessed in 836 randomly recruited participants from an urban developing community of African ancestry. RESULTS Log CRP was strongly correlated with age, steady-state pressures (indexed by mean arterial pressure (MAP)), indexes of adiposity, central systolic BP (SBPc), central pulse pressure (PPc), the augmented (reflected) (AP) and forward (P1) pressure wave components of PPc, aortic pulse wave velocity (PWV), and the effective reflecting distance (ERD) (P < 0.0001 for all). Moreover, before adjustments SBPc, PPc, AP, P1, PWV, and ERD increased incrementally across low, moderate and high risk categories of CRP (P < 0.0001). However, in multivariate models with adjustments for confounders, no independent association between CRP and SBPc, PPc, AP, P1, PWV, or ERD was noted (P > 0.10 for all) and multivariate adjusted central hemodynamic variables were similar in participants with low, moderate, or high risk CRP concentrations. The lack of independent relationship between CRP concentrations and central aortic hemodynamics was reproduced in normotensives, hypertensives, men, women, lean and overweight/obese participants, and in participants lesser or ≥ 60 years of age. CONCLUSION Even in communities with prevalent high CRP concentrations, low-grade inflammation as indexed by CRP, does not independently account for increases in aortic PP, the component waves, or the determinants of PPc.

Relationship of predominantly mild current smoking to out-of-office blood pressure in a community sample in Africa

Objectives As the impact of mild smoking on blood pressure (BP) is uncertain, we assessed the relationship between predominantly mild current smoking and out-of-office BP and the effect of angiotensin-converting enzyme (ACE) genotype on this relationship in a community sample of black African ancestry. Methods In 689 participants randomly recruited from an urban, developing community of black African descent, we assessed smoking habits, out-of-office (24-h), and in-office conventional and central (applanation tonometry) BP, and ACE insertion (I)/deletion (D) variant genotype. Results A total of 14.5% (n = 100) were current smokers, the majority being mild (72%, 7.4 ± 4.6 cigarettes/day). Despite current smokers having only modest increases in in-office (P < 0.05) and similar central aortic BP values as nonsmokers, current smokers had higher unadjusted (P < 0.005–P < 0.0005) and multivariate adjusted 24-h SBP/DBP (mmHg; smokers = 123 ± 15/76 ± 10; nonsmokers = 118 ± 14/72 ± 9; P < 0.005–P < 0.0005) than nonsmokers, effects that were DD genotype-dependent (P < 0.005 for interaction) and replicated in sex-specific groups, nondrinkers, and in overweight and obese. Current smoking was second only to age in the quantitative impact on 24-h DBP. Smoking 4.6 cigarettes per day (one standard deviation) translated into increases in 24-h SBP (mmHg) of 2.12 [confidence interval (CI) = 1.77–2.47] in all participants and 3.62 (CI = 3.13–4.12) in participants with the DD genotype. The risk of uncontrolled 24-h BP was increased in smokers as compared to nonsmokers (adjusted odds ratio = 1.87, CI = 1.02–3.41, P < 0.05), an effect that was enhanced in participants with the DD genotype (adjusted odds ratio = 4.01, CI = 1.59–10.09, P < 0.005). Conclusion Mild current smoking is independently associated with an appreciable proportion of out-of-office BP in a black African community, an effect that is ACE genotype-dependent.

Intra-familial aggregation and heritability of aortic versus brachial pulse pressure after imputing pretreatment values in a community of African ancestry.

Aim: To compare the intra-familial aggregation and heritability of central (aortic) (PPc) versus peripheral (brachial) (PPp) pulse pressure after imputing pretreatment blood pressures (BPs) in treated participants in a community of black African ancestry. Methods: Central PPc [generalized transfer function (GTF) and radial P2-derived] was determined with applanation tonometry at the radial artery (SphygmoCor software) in 946 participants from 258 families with 23 families including three generations from an urban developing community of black Africans. In the 24.1% of participants receiving antihypertensive treatment, pretreatment brachial BP was imputed from published overall averaged effects of therapy grouped by class and dose, specific for groups of black African descent. From these data PPc was estimated from proportionate differences in central aortic and brachial PP. Heritability estimates were determined from SAGE software. Echocardiography was evaluated in 507 participants in order to determine stroke volume. Results: With adjustments for confounders, parent–child (P < 0.05) and sibling–sibling (P < 0.0005) correlations were noted for log PPc, whilst for log PPp only sibling–sibling correlations were noted. No mother–father correlations were noted for either PPc or PPp. Independent of confounders the heritability for log GTF-derived (h2 = 0.33 ± 0.07, P < 0.0001) and P2-derived (h2 = 0.30 ± 0.07, P < 0.0001) PPc was greater than the heritability for log PPp (h2 = 0.11 ± 0.06, P < 0.05) (P < 0.05 for comparison of heritability estimates). Conclusion: After imputing pretreatment BP values, central aortic PP is significantly more inherited than brachial PP. These data suggest that in groups of African descent the genetic determinants of PP may be underestimated when employing brachial rather than central aortic PP measurements.

Intrafamilial Aggregation and Heritability of Left Ventricular Geometric Remodeling Is Independent of Cardiac Mass in Families of African Ancestry

BACKGROUND Whether left ventricular (LV) geometric remodeling, as indexed by relative wall thickness (RWT), aggregates in families and is inherited independent of LV mass (LVM) and additional confounders is uncertain. METHODS We determined whether RWT as assessed from 2D targeted M-mode echocardiography shows intrafamilial aggregation and heritability independent of LVM in 181 nuclear families (73 spouse pairs, 403 parent-child pairs, and 177 sibling-sibling pairs) with 16 families including 3 generations from an urban developing community of black Africans. Intrafamilial aggregation and heritability estimates (S.A.G.E. software) were assessed independent of confounders, including central aortic systolic blood pressure (SBPc) (radial applanation tonometry and SphygmoCor software). RESULTS Independent of confounders including SBPc, LV RWT was correlated in parent-child (r = 0.32, P < 0.0001) and sibling-sibling (r = 0.29, P < 0.0001), but not in spouse (r = 0.11, P = 0.33) pairs. The relationships between parent-child (r = 0.28, P < 0.0001) and sibling-sibling (r = 0.24, P < 0.001) pairs persisted with further adjustments for LVM or LVM indexed to height(2.7) (LVMI). Similarly, independent of confounders, LV RWT showed significant heritability (h(2) ± SEM = 0.56 ± 0.09, P < 0.0001) and this persisted with further adjustments for LVM (h(2) ± SEM = 0.48 ± 0.09, P < 0.0001) or LVMI (h(2) ± SEM = 0.49 ± 0.09, P < 0.0001). CONCLUSIONS In a group of African ancestry, independent of LVM, LV geometric remodeling shows significant intrafamilial aggregation and heritability. Genetic factors may in-part determine the LV geometric remodeling process independent of the extent of cardiac hypertrophy.

Intrafamilial aggregation and heritability of office-day blood pressure difference in a community of African ancestry: implications for genetic association studies.

ObjectiveAn inability to show consistent relationships between gene variants and blood pressure (BP) may be confounded by the use of office BP measurement. Whether the difference between office BP and day BP (office−day) is genetically predetermined is unknown. We therefore aimed to determine the intrafamilial aggregation and heritability of office−day BP. Patients and methodsNurse-derived office BP (mean of 5 measurements according to guidelines) and 24-h ambulatory BP were determined for 592 participants from 198 families (67 spouse pairs, 361 parent–child pairs, and 169 sibling–sibling pairs), with 12 families having three generations, from an urban developing community of black Africans. Heritability estimates were determined using SAGE software. ResultsWith adjustments for confounders, office systolic BP (SBP) (h2=0.35±0.09, P<0.0001) showed comparable heritability estimates to 24-h SBP (h2=0.33±0.09, P<0.0001). Similarly, with adjustments for confounders, office diastolic BP (DBP) (h2=0.37±0.09, P<0.0001) showed comparable heritability estimates as 24-h DBP (h2=0.35±0.09, P<0.0001). However, multivariate adjusted heritability estimates of day SBP (h2=0.29±0.09, P<0.0001) and DBP (h2=0.33±0.09, P<0.0001) were not diminished by further adjustments for office SBP (h2=0.42±0.09, P<0.0001) or DBP (h2=0.34±0.09, P<0.0001). Further, independent of confounders, office−day BP showed significant intrafamilial aggregation and heritability (SBP: h2=0.51±0.10, P<0.0001; DBP: h2=0.37±0.09, P<0.0001), effects that persisted with further adjustments for office, day, or day−night BP (P<0.0005 for SBP and DBP). ConclusionAlthough office and ambulatory BP may show similar heritability estimates, genetic associations with carefully determined office BP measurements may be confounded by the heritability of office−day BP differences.