43 / 68
CARDIOVASCULAR JOURNAL OF AFRICA • Volume 30, No 3, May/June 2019
AFRICA
171
Patients’ median aspartate aminotransferase (AST) (
p
=
0.006) and alanine aminotransferase (ALT) (
p
=
0.009) levels
were higher than those with non-valvular AF. Other laboratory
data, including haemoglobin, were comparable between the
two groups. Those with valvular AF had a significantly lower
HAS-BLED score (
p
<
0.001) than the non-valvular AF patients.
An anticoagulation prescription was significantly more frequent
in patients with valvular AF than the non-valvular AF patients
(
p
=
0.005). Few patients with non-valvular AF were on a
non‐vitamin K antagonist (NOAC). The prescriptions of other
medications were comparable between the two groups.
There was a total of 20 (14.5%) deaths during the 12-month
follow-up period. There were 13 (13.8%) and seven (15.9%)
deaths in patients with valvular and non-valvular AF, respectively
(
p
=
0.746).
Discussion
Only a few studies have been done on AF in SSA, despite its
growing public health importance. This study describes the
characteristics and outcomes of patients with AF in a developing
SSA country.
Hypertension, heart failure and rheumatic heart disease
(RHD) were the most common underlying diseases for AF in
our relatively young cohort. We also identified a low prevalence
of diabetes mellitus, cigarette smoking, alcohol consumption
and chronic kidney disease among patients. There was a baseline
difference in age and burden of cardiovascular disease in patients
with valvular AF and those with non-valvular AF. However,
despite these differences, death rates (all-cause mortality) were
comparable between the two groups. The overall mortality rate
during the 12-month follow-up period was high (14.5%).
Consistent with previous reports from SSA, our cohort
was younger (67 years) than patients in developed countries,
where most cases are recorded around the eighth decade of
life.
7,23-26
The mean age of African patients with AF has, on
average, been a decade younger than patients in developed
countries.
24,25
A shorter life expectancy in SSA and differences in
the epidemiology of diseases associated with AF in developed
and developing countries partly explain this age difference.
4
The
main risk factors of AF in SSA (RHD, hypertension and heart
failure) tend to present at a younger age and in a more severe
form than in developed countries.
27-30
The finding of female predominance in the present study is
consistent with previous results in SouthAfrica andCameroun.
24,25
This pattern is likely explained by gender-based differences in
potential underlying diseases (particularly RHD). Although
there is no apparent gender predilection for acute rheumatic
fever, RHD tends to be more common in females than males.
27,31
Also, there is a two-fold increased risk for AF in women with
valvular heart disease compared to their male counterparts.
1
Consequently, the proportion of female patients among our
valvular AF patients was significantly higher compared to
patients with non-valvular AF.
In the present study, RHD was found in more than a third
(37.7%) of participants. This is not surprising as RHD is
much more common in patients with AF in Africa than in the
developed world.
7
In the global Randomized Evaluation of
Long-Term Anticoagulation Therapy (RE-LY) AF registry that
enrolled patients presenting with AF to emergency departments,
RHD was present in only 2.2% of North American patients
compared to 21.5% in Africa.
7
Likewise, the proportion of
African patients with RHD in South Africa and Cameroun
was reported as 21 and 25.6%, respectively.
24,25
Although the
exact reasons for the higher frequency of RHD in our cohort
than in other African settings are unclear, one may be that
there is variation in the overall burden of RHD across African
countries. About a third (32%) of our cohort was found to have
moderate-to-severe mitral stenosis or prosthetic heart valve AF
and was classified as having valvular AF.
21,22
As a consequence
of the higher prevalence of RHD, valvular AF patients were
significantly younger and more likely to be female than those
with non-valvular AF. This finding has clinical implications, as
the presence of AF in the reproductive years may be associated
with worse pregnancy outcomes.
We found a high burden of hypertension and heart failure
among our participants, consistent with previous AF studies
worldwide.
7,8,24,25
Unfortunately, these diseases are likely to remain
undiagnosed or poorly controlled inAfrica compared to developed
countries.
7,32
Early detection and control of hypertension have the
potential to reduce the occurrence and progression of AF in
SSA. In contrast to reports from developed countries, diabetes
and coronary artery diseases were infrequent in our cohort, as in
other African studies.
7,8,24,25
Likewise, in the RE-LY AF registry,
the prevalence of diabetes and coronary artery diseases were the
lowest among Africans compared to other regions of the world.
7
With the increasing burden of non-communicable diseases in
SSA, the contribution of diabetes and coronary artery disease to
the overall prevalence of AF in this region is likely to rise.
33
Comparable to previous studies in Africa, our patients with
non-valvular AF were more likely to have hypertension than
those with valvular AF.
7,26
Also, chronic kidney disease, obesity
and cigarette smoking were more frequent in this group. The
presence of these co-morbidities is known to adversely affect
survival in individuals with AF.
34
The observed higher burden of
co-morbidities in non-valvular AF patients compared to valvular
AF may in part be explained by the older age of participants in
this group. This burden of co-morbidities partly reflects in high
thromboembolic risk (mean CHA
2
DS
2
-VASc 3.6), comparable to
the risk previously reported in the European GARFIELD study
(mean CHA
2
DS
2
-VASc 3.3). As previous studies in Africa have
stratified patients’ thromboembolic risk using CHADS
2
scores,
it was difficult to compare our findings to other SSA countries.
Using the CHADS
2
score, AF patients in SSA have consistently
been reported to have low thromboembolic risk.
7,24,25
Similar to the data obtained previously in Kenya, a significant
proportion of our patients had a history of stroke/TIA.
26
The
burden of stroke/TIA in our AF cohort was, however, higher
than seen in previous reports from other countries in SSA.
7,24,25
Although the majority of patients with stroke/TIA were those
with non-valvular AF, the burden of stroke/TIA (9.3%) in the
young cohort with valvular AF was still higher than what was
previously reported in South Africa (2.4%). The reason for this
disparity is unclear but may lead to speculation about variation
in thromboembolic risks across communities.
The majority of our patients with non-valvular AF had
CHA
2
DS
2
-VASc scores
≥
2 and required oral anticoagulation
to prevent stroke. A higher proportion of our AF patients were
on warfarin anticoagulation compared to other reports across
the globe.
7
Although NOACs are indicated as an alternative