CARDIOVASCULAR JOURNAL OF AFRICA • Vol 23, No 9, October 2012
AFRICA
481
overall prevalence of RVDD was higher than that of RVSD, and
the highest prevalence of the latter was recorded in subjects with
EH, while prevalence of RVDD was high across the groups.
The study has also described the determinants and correlates of
RVSD and RVDD.
Several studies have previously reported that RV disease
develops in parallel with a similar process on the left side
among hypertensive patients,
1,2
likely as a result of ventricular
interdependence. Ventricular interdependence is defined as the
forces that are transmitted directly from one ventricle to the
other through the myocardium and pericardium, independent
of neural, humoral or circulatory effects. It is a consequence
of the close anatomical association between the ventricles: the
ventricles are encircled by common muscle fibres, share a septal
wall, and are enclosed within the pericardium.
20
In agreement with this hypothesis, LVEF, the index for LV
systolic function, was the only independent determinant of
RVSD. In addition, the present study also recorded the highest
prevalence of RVSD among subjects with EH who had the lowest
mean LVEF and the worst LV systolic function, while the lowest
prevalence of RVSD was recorded among those with NG, who
correspondingly had the highest mean LVEF.
The prevalence of RVSD in the present study (32.03%),
determined using both TAPSE and S
m
,
was only slightly higher
than what we reported previously (29.06%) using TAPSE alone.
1
This suggests that the inclusion of S
m
in the determination of
RVSD contributes very little over that of using TAPSE alone.
This finding therefore supports the use of TAPSE alone to
determine RVSD in hypertensive subjects.
The prevalence of RVSD in the present study was lower than
what was reported by Puwanant
et al
.
(58%)
among hypertensive
subjects using reduced TAPSE (
<
15
mm) alone, perhaps because
51%
of the patients in their series had coronary artery disease,
37%
had diabetes mellitus and 32.5% had cardiomyopathies.
21
In addition, the patients in the latter study were older than ours
(
mean age 72
±
14
vs 51
±
14
years), and we have previously
shown that older age is significantly associated with reduced
TAPSE.
1
Furthermore, differences in the aetiology of heart
disease in the two studies could have amplified the disparities
in RVSD.
21
The prevalence of RVDD was high across the LV geometric
groups, without significant statistical differences. Therefore, the
pattern of RVDD in the studied population was not similar to
that of RVSD. In the present study, subjects with NG, who had
a mean mitral filling E:A ratio of 1.17 and the lowest mean age
and RVOTd, turned out to have (albeit non-significantly) the
highest prevalence of RVDD. We found that age was the only
determinant of RVDD (
p
=
0.023),
and correlated significantly
and negatively with the lateral tricuspid annular E
m
:
A
m
ratio; the
index of RVDD (
r
=
–0.237;
p
=
0.016).
The relatively lower
mean age of patients with NG could therefore have influenced
the observed high prevalence of RVDD in them, in view of the
negative correlation between age and the E
m
:
A
m
ratio.
Innelli
et al
.
recently reported that the E
m
:
A
m
ratio consistently
and progressively decreased with age in an apparently healthy
population, from 1.9
±
0.8
in the 10–19-year-olds to 0.95
±
0.3
in
50–59-
year-old age group.
22
It then appears that an E
m
:
A
m
ratio of
<
1.0
may not be reliable for the assessment of RVDD in a young
population, such as those in the NG group. The dependence
of RVDD on age could be attributed to the increase in arterial
stiffness of the pulmonary vessels with ageing.
23
Another possible explanation for the high prevalence of
RVDD across the groups in our study could be the fact that
RVDD in hypertensives has been shown to develop early, before
apparent systolic dysfunction, and before RV dilatation or RV
hypertrophy.
17
This implies that RVDD could be the earliest index
of RV affectation in hypertension, and perhaps occurs before LV
geometry becomes abnormal. Therefore, the E
m
:
A
m
ratio could
potentially be used as a high-sensitivity screening tool for RV
disease, but taking the age of the individual into consideration.
In contrast to our finding, Cicala
et al.
previously reported
that mitral annular E
m
:
A
m
ratio and body mass index were the
only predictors of RVDD, while age, DBP, heart rate, septal
and RV wall thickness were not associated with the RVDD.
2
We
found no relationship between RVDD and LVMI, but did not
assess the relationship between RVDD and mitral annular or
mitral valve filling variables in the present study.
Several conditions have been associatedwithRVDD, including
both RV pressure and volume overload pathologies, primary lung
disease, IHD, congenital heart disease, cardiomyopathies, LV
dysfunction (via ventricular interdependence), systemic diseases
and the physiological aging process.
17
IHD is still uncommon in
sub-Saharan Africa, previously found among 8.7% of subjects
referred for echocardiography in Kano, Nigeria, and among
10.4%
of the Soweto community in South Africa.
9,24
We excluded
subjects with IHD, however, from the present study.
The limitations of the present study include the use of E
m
:
A
m
ratio alone to assess RVDD. However, this index is one of the
recommended indices approved by the American Society of
Echocardiography for assessing RVDD, and has been shown to
represent global RV diastolic function.
2,17
Secondly, magnetic
resonance imaging appears to be superior to echocardiography
and other techniques in studying the right ventricle.
25
However,
echocardiography has acceptable sensitivity, and is widely
available and affordable, and therefore has an important role in
studying the right ventricle, despite its limitations.
Another limitation was the use of Teicholz’s formula to
estimate LVEF, which has the inherent tendency to over-estimate
it in the presence of abnormal LV geometry.
14
To minimise the
inaccuracy of the Teichholz’s formula, subjects with regional
wall motion abnormality were excluded from the study. The
formulae that estimate LV mass using measurements obtained
from two-dimensional guided M-mode echocardiography
have several limitations, including sub-optimal accuracy in
the presence of abnormal LV geometry, large inter-observer
variability and poor inter-study reproducibility.
14,16
To minimise
this, about 50% of the echocardiograms were carried out by
the principal author (KMK), while the remaining 50% were
carried out by the co-authors (HS and MNS). We are presently
in the process of estimating the inter-observer variability for our
echocardiography laboratory.
Conclusion
This study has described, perhaps for the first time, the pattern
of RVSD and RVDD in hypertensive subjects grouped according
to pattern of LV geometry. The majority of subjects (42.2%) had
NG, while 29.7% had EH, 22.7% had CH, and 5.5% had CR. The
highest prevalence of RVSD was recorded in subjects with EH
whohad the lowestmeanLVEFand theworstLVsystolic function,
