Determinants of left ventricular hypertrophy and its regression in people of African ancestry in South Africa
Date
2008-07-10T09:04:51Z
Authors
Libhaber, Elena Neustadt
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
ABSTRACT
There is substantial evidence to suggest that independent of conventional BP, LV
mass (LVM) is higher in African-Americans than in European-Americans a difference that
may translate into a higher prevalence of cardiovascular diseases. In the present thesis I
assessed whether LVM is similarly elevated in groups of African descent living in Africa,
and subsequently whether 24-hour, day or night BP or indices of arterial stiffness could
explain the variability in LVM beyond conventional BP in this population group. As there
is considerable controversy as to whether 24-hour BP measurements are better
predictors of the regression of LVH than conventional BP and whether antihypertensive
agents that target the renin-angiotensin system (RAS) regress LV hypertrophy (LVH)
independent of BP in groups of African descent, in the present thesis I therefore also
assessed these questions.
In 141 healthy adult participants obtained from a random sample of nuclear
families (n=399) of African ancestry living in Soweto, I determined that LVM adjusted for
body surface area to the first power was an appropriate allometric signal to account for growth effects on LVM. The allometric signals established in other populations
considerably over-adjusted for LVM in the group that I studied with marked negative
relations noted. After adjusting for body surface area I noted upper thresholds of LVM
index (LVMI) of 134 g/m2 for men and 112 g/m2 for women. As compared to thresholds
described for other population samples these thresholds were noted to be modestly
higher.
In 187 women from randomly recruited nuclear families of African ancestry, after
appropriate adjustments, conventional BP was as closely associated with LVMI as 24-
hour BP, and daytime BP was as closely associated with LVMI as night-time BP in
women. However, in 110 men from randomly recruited nuclear families of African
ancestry, after appropriate adjustments, only night-time BP was associated with LVMI,
an effect that was independent of conventional BP (r=0.21, p<0.05). Indices of nocturnal
decreases in BP were not associated with LVMI in either gender group. Furthermore, in randomly recruited nuclear families of African ancestry, after appropriate adjustments,
including systolic BP or pulse pressure, pulse wave velocity (an index of arterial stiffness
assessed using applanation tonometry) was independently associated with LVMI in
women (n=204, r=0.25, p<0.0005), but not in men (n=123, r=-0.07).
In 173 hypertensive patients of African descent of whom 64 were previously
untreated and 109 were previously treated, I assessed whether ambulatory BP is a
better predictor of on-treatment decreases in LVMI over a 4 month treatment period. In
the previously untreated patients, the regression in LVMI correlated to a similar degree
(p<0.09) with decreases in conventional (r=0.34; p<0.005) and 24-hour (r=0.26; p<0.04)
systolic BP. In this same study sample followed prospectively for 25 months, accounting
for effects on ambulatory BP at each time point, the use of the angiotensin-converting
enzyme inhibitor, enalapril, was not associated with LVMI, whereas, on-treatment
conventional systolic BP (p=0.01) and night-time systolic BP (p=0.01) were associated
with LVMI.
In a further study conducted in 87 patients of African ancestry with hypertension
and LVH, I showed that changes in systolic ambulatory BP (daytime, r=0.46, p=0.006)
were predictive of changes in LVMI after 2 months of treatment with an angiotensin II
receptor blocker (candesartan), ACE-I (ramipril) and the diuretic agent,
hydrochlorothiazide. Moreover, in a final study I showed that in hypertensive patients of
African ancestry, initiating therapy with the diuretic, indapamide SR and then adding the
ACE-I, perindopril 4 mg (n=42), was equally as effective as amlodipine (calcium channel
blocker) (n=44) therapy at reducing ambulatory BP and LVMI.
Thus, in conclusion, groups of African descent living in Africa have only
marginally higher thresholds for LVM than other population groups. Moreover, in this
population group, nocturnal BP has a conventional BP-independent effect on LVMI in
men, but not in women, whereas arterial stiffness has a conventional BP-independent
effect on LVMI in women, but not in men. Further, in this population, reductions in LVM
produced by antihypertensive therapy appear to be equally as closely related to conventional as ambulatory BP and in contrast to findings in groups of European
ancestry, where RAS blockers produce unique benefits on LVM beyond conventional BP
reductions, in groups of African ancestry in Africa, RAS blockers produce no BPindependent
reductions in LVM. Moreover, in this population, decreases in LVM in
patients with LVH produced by RAS blockers are related to ambulatory BP changes and
despite the ineffectiveness of RAS blockers on BP when used as monotherapy in this
population, RAS blockers together with diuretics are equally as effective in decreasing
BP and LVM as compared to a class of antihypertensive agents with established efficacy
(calcium channel blockers). Hence when compelling indications for RAS blockade exist,
RAS blocker-diuretic combinations are effective therapy in patients of African descent
living in Africa.
Description
Keywords
left ventricular hypertrophy, African ancestry