CARDIOVASCULAR JOURNAL OF AFRICA • Vol 23, No 6, July 2012
AFRICA
315
When multivariate regression analysis was applied to all the
variables, including risk factors for CAD and SPECT myocardial
perfusion indices, only age emerged as a significant factor (Table
4). In addition, univariate analysis showed that patients with
elevated homocysteine levels had a lower BMI (26.6 kg/m
2
)
compared to those with normal homocysteine levels (30.6 kg/
m
2
) (
p
=
0.002).
When homocysteine levels were log transformed and the
correlation was determined with the myocardial indices, no
significant correlation was found for SSS, SRS, SESV and SEF
over the whole population, and sub-group analysis found no
correlation with various known risk factors (Table 4). We did
note however, a significant correlation between homocysteine
level and SSS in the patients between 50 and 64 years.
The extent of disease was important. A significant correlation
was noted between the extent of disease measured by the number
of involved coronary artery territories demonstrating infarct,
as seen by the SRS evaluated from the myocardial perfusion
scintigraphy, and the plasma homocysteine level, where the
presence of disease in two or three coronary artery territories
was significant (Table 5, Figs 1, 2). Such a correlation was not
as strong when comparing the extent of ischaemia as seen by the
SSS, with a raised homocysteine level, where there was only a
significant correlation in single-vessel disease.
Discussion
This study examined the relationship between plasma
homocysteine levels and myocardial SPECT perfusion indices.
Previous investigators have reported a good correlation between
scintigraphically detected ischaemia and plasma homocysteine
levels in patients on haemodialysis.
14
A further study demonstrated elevated homocysteine levels
with reduced regional left ventricular ejection fraction in an
asymptomatic group of young patients,
1
while others found a
high prevalence of patients with elevated homocysteine levels
in patients with heart failure with preserved ejection fraction.
15
This was similar to the figures we obtained where the presence
of infarct seemed to have more influence on homocysteine levels
than ischaemia.
Another study demonstrated the association of homocysteine
with left ventricular dysfunction independent of the presence of
CAD.The same study did not demonstrate a significant correlation
with myocardial ischaemia.
16
These results, together with the
failure of homocysteine-lowering therapy to reduce cardiac
events, show the relationship between raised homocysteine levels
and myocardial damage remains poorly understood,
17-19
although
there is clearly a need to investigate this link.
20,21
Myocardial SPECT imaging analysed with software such as
4D gives semi-quantitative indices for assessment of myocardial
perfusion, in addition to its ability to assess left ventricular
function
22,23
and regional wall abnormalities.
24
When we
subdivided our study patients based on the number of affected
coronary artery territories, we noted a significant positive but
weak correlation between SSS and patients with dual- or triple-
vessel CAD, but a stronger link with the SRS, again suggesting
infarct was more important than ischaemia. In addition to this
finding, we also demonstrated a significant positive correlation
between SRS and plasma homocysteine level in patients with
single-vessel CAD.
These findings suggest that higher homocysteine levels are
associated with more severe and extensive infarct. This may
help to explain why such patients with established, irreversible
disease are unlikely to benefit from reduction in homocysteine
levels alone. This is somewhat confirmed by the findings in
our study that higher SESV and lower SEF were observed in
TABLE 2. NUMBER OF RISK FACTORS PER PATIENT
SHOWINGA SIGNIFICANT CORRELATION BETWEEN THE
NUMBER OF RISK FACTORS PER PATIENT
AND RAISED HOMOCYSTEINE LEVELS
Number of risk
factors for CAD
present in each
Patients with the
number of risk
factors in column 1
Patients with raised
homocysteine
n
%
1
29
5
14
2
24
8
33
3
48
17
35
4
18
11
61
CAD
=
coronary artery disease,
p
=
0.028 (
χ
2
).
TABLE 3. MEAN MYOCARDIAL PERFUSION INDICES
IN PATIENTSWITH ELEVATEDAND NORMAL
HOMOCYSTEINE LEVELS
Index of
myocardial function
Elevated
homocysteine level
Normal
homocysteine level
p
SSS
11.3
6.9
0.02
SRS
3.4
2.1
ns
SEF (%)
54
64
0.02
SESV (ml)
137
105
0.03
SSS
=
summed stress score, SRS
=
summed rest score, SEF
=
post-stress
left ventricular ejection fraction, SESV
=
stress end-systolic volume.
TABLE 4. MULTIVARIATE REGRESSIONANALYSIS FOR
RISK FACTORSAND MYOCARDIAL PERFUSION INDICES
AND PLASMA HOMOCYSTEINE LEVELS
Factor
Odds ratio
SE
Z
p
>
[Z]
Age
1.076
0.033
2.36
0.019
Diabetes mellitus
1.104
0.681
0.16
ns
Hypertension
3.635
3.280
1.43
ns
Dyslipidaemia
1.209
0.705
0.33
ns
Smoker
3.525
2.771
1.52
ns
SSS
1.002
0.036
0.05
ns
SRS
1.140
0.085
1.75
ns
SEF
0.969
0.029
–1.05
ns
SESV
0.999
0.021
–0.06
ns
SSS
=
summed stress score, SRS
=
summed rest score, SEF
=
post-stress
left ventricular ejection fraction, SESV
=
stress end systolic volume,
ns
=
not significant.
TABLE 5. CORRELATION BETWEEN ELEVATED
HOMOCYSTEINEAND THE SSSAND SRS IN PATIENTS
WITHA GIVEN NUMBER OF DISEASED CORONARY
ARTERY TERRITORIES, AS SEEN ON MPS
Number of abnormal
coronary artery territories
SSS
SRS
r
p
r
p
0
–0.05
ns
–0.1
ns
1
0.31 0.005
0.1
ns
2
–0.05
ns
0.18
0.049
3
0.15
ns
0.28
<
0.001
SSS
=
summed stress score, SRS
=
summed rest score,
r
=
correlation
co-efficient,
p
=
significance level, ns
=
not significant.
