CARDIOVASCULAR JOURNAL OF AFRICA • Vol 22, No 3, May/June 2011
138
AFRICA
from cells in the arterial wall,
26
and thereby maintaining cell
cholesterol homeostasis, HDL-C is also antithrombotic,
27
and
possesses antioxidant and anti-inflammatory properties.
28,29
Our
HIV-infected participants had lower serum HDL-C levels and
were 3.7 times more likely to have low HDL-C levels than their
matched controls, while the lowest levels were seen in the nadir
CD
4
cell count group. The protective effect of HDL-C was lost
in our HIV-infected group due to the low levels.
Furthermore, Patel
et al
. demonstrated that triglycerides
contribute to atherosclerosis-mediated inflammation by their
direct effect on the endothelium and also potentially by attenu-
ating the protective effects of HDL-C against vascular inflam-
mation.
30
It could therefore be expected that the low serum
HDL-C and high TG levels and higher TG/HDL-C ratio seen
in the HIV-infected versus uninfected participants worsened the
inflammation, as reflected by the higher levels of and odds ratios
for having higher levels of hsIL-6 and hsCRP in the infected
group. The inverse correlation of HDL-C with hsIL-6 levels (
r
=
–0.21,
p
=
0.001), only seen in our HIV-infected group, further
emphasises the inflammation in this group. Clinical and epide-
miological studies showed that HDL-C concentration is often
inversely related with plasma levels of cytokines in atheroscle-
rotic cardiovascular diseases.
31
While IL-6 is an early stimulator of the inflammatory process
and CRP is produced in response to IL-6 secretion, CRP is
thought to induce ICAM and VCAM secretion.
32
These adhesion
molecules are known to be expressed in arteries
in vivo
at sites
of developing atherosclerosis
33
and indicate vascular endothelial
injury and dysfunction.
3,20,34
The CRP-induced expression of
endothelial adhesion molecules is inhibited by HDL-C.
31
The
HIV-infected participants in our study showed lower HDL-C and
higher hsIL-6, hsCRP, sICAM-1 and sVCAM-1 levels compared
to their uninfected controls. The inflammatory process was
thereby clearly activated, resulting in endothelial injury. The
HIV-infected participants were two and four times more likely to
express higher levels of sICAM-1 and sVCAM-1, respectively,
further indicating endothelial injury in this group.
It was found that HIV-1 Tat protein (therefore HIV itself)
induced the expression of ICAM-1 and VCAM-1 and this could
be a possible mechanism by which HIV-1 infection contributes
to endothelial injury and accelerated atherosclerosis.
35,36
It could
therefore be expected that the HIV-infected, never antiretroviral-
treated participants with definite signs of endothelial injury
would also show signs of endothelial dysfunction.
Endothelial dysfunction results in increased arterial stiff-
ness,
37
which progresses as arteries become more damaged.
38
Using femoral PWV, previous studies showed increased aortic
(central) stiffness in treated
39
and untreated
9
HIV-infected indi-
viduals. In our study, the peripheral cr-PWV did not differ
between the HIV-infected participants and their age-, gender-,
BMI- and locality-matched controls. However, age, although
weak, correlated positively with carotid radial PWV (
r
=
0.14,
p
=
0.01) in the HIV-infected group alone.
Age is one of the principal factors modulating PWV,
40
which is
greatest in the elastic aorta and least in muscular arteries such as
those of the upper limb.
41
When divided into 10-year-interval age
groups, our HIV-infected participants showed a positive trend of
increasing peripheral cr-PWV with age. Since changes in PWV
in muscular arteries are not normally found with increasing age,
42
as in the uninfected group of the present study, this increase in
PWV in muscular arteries of the HIV-infected participants was
a significant result.
Although increased stiffness could indicate atherosclerosis
per se
, over and above that due to aging,
43
our results being in
muscular arteries may point to premature vascular aging in the
HIV-infected participants. This is in agreement with the results
of Lorenz
et al.
who obtained a higher ‘vascular’ age of four
to five years for HIV-infected patients (treated) compared to
controls.
13
Our results indicate that HIV-1 without the effect of
treatment might contribute to accelerated vascular aging and
possible early atherosclerosis.
Endothelial dysfunction is also associated with a prothrom-
botic state,
44
and studies which investigated the prothrombotic
state in HIV-infected populations showed increased levels of
PAI-1 activity and fibrinogen.
4,45,46
However, in our study neither
the PAI-1 activity nor fibrinogen levels were increased in the
HIV-infected subjects, indicating no signs of a prothrombotic
state. Our findings are in agreement with the findings of James
et al.
, who found that HIV infection was not associated with the
fibrinogen concentration in Africans.
47
It is known that black
South Africans of African ancestry have high levels of fibrino-
gen,
48
and it may therefore be that ethnic effects on plasma fibrin-
ogen may have masked the potential effect of HIV infection.
This study has limitations and strengths. It had a case–control
design and control participants were carefully matched to the
infected participants according to age, gender, BMI and local-
ity. When viewing previous studies regarding HIV and cardio-
vascular risk, our study population was unique as they were
unaware of their infection status and therefore had never received
antiretroviral treatment. Therefore, although the evidence of no
self-reported diseases was not evaluated, the differences found
could probably be attributed to the infection itself and not ARV
treatment. Furthermore, although South Africa has the highest
infection rate in the world,
7
data on cardiovascular changes and
risk in HIV-1-infected South Africans are limited.
A limitation of the study was that unfortunately, probably due
to stigmatisation, which still exists among South African individ-
uals
49,50
and other aspects such as the illness itself and poverty,
51
the participants did not visit the local clinic or hospital for follow
up and CD
4
cell count determinations. Therefore the sample
size of the nadir CD
4
cell count group was very small and those
results should be interpreted with caution. Also, the subjects in
this study were newly identified as being HIV infected and there-
fore the duration of the infection was not known.
The lack of increase in fibrinogen, PAI-1 and cr-PWV levels
in the infected group may be related to the possible short dura-
tion of the infection, as is also speculated in the study of James
et al.
47
This is confirmed by the tendency of increased (although
not statistically significant) fibrinogen levels in the nadir CD
4
cell count group.
A longitudinal study is therefore proposed to further investi-
gate the influence of HIV on the endothelium and prothrombotic
state of Africans. A recommendation for future studies would
be to perform carotid intima–media thickness measurements to
verify endothelial damage and probable atherosclerosis.
Conclusion
Our findings suggest inflammatory injury of the endothelium,
pointing to endothelial dysfunction of never antiretroviral-treat-
