Does measurement of apalipoproteins add to
the estimate of coronary artery disease risk?
Gunnar Sigurdsson, M.D., Ph.D. Department of Medicine, Reykjavik City Hospital, leelandie Heart
Association, Reykjavik, leeland
Introduction
Apolipoproteins, the major non-enzymatic protein
constituents of lipoproteins, play an important role
in determining the function and fate oflipoproteins.
They influence both the structure of lipoproteins
and their fate by either facililating or inhibiting the
uptake of lipoproteins through specific receptors or
by enzyme activation.
A few reports have been published indicating
that measurements of apalipoproteins maybe
valuable in predieting future coronary artery disease
risk , even more so than lipoprotein lipid
measurements. [1 -5]. However, most ofthe studies
have been retrospective in nature with all the inhe–
rent limitationsofthat approach. The Reykjavik
Study, a prospective health survey carried out by
the leelandie Heart Association since 1967, gave us
an opportunity to test this hypothesis.
Study population and methods
Partidpants in this prospective health survey (stage
IV of the Reykjavik Study) were male residents of
the dty's area, randomly selected from the Natio–
nal Register in various age groups, bom between
1908 and 1933; the mean age at baseli ne was 56
years. Of 2060 men contacted 1444 participated
(70%). However, only the frozen serumsamples
available from 1332 partidpants were included in
this survey which was carried out from November
1979 to March 1981.
Serum cholesterol and triglycerides were
measured at baseline and additional serum samples
kept in sealed plastic caps at -20°C and then thawed
(once) eight years later and analysed for apo (a),
apo-B and apo-A-1. Apo (a) was determined with
a 2-site immunoradiometric assay, using two diffe–
rent monoclonal antibodles in excess (Pharmada
Klinisk kemi
i
Norden 2, 1994
DiagnosticNB, Uppsala). ThestudyofJauhiainen
et al [6] showed that freezing at -20°C for severel
years did not affect the stability of apo (a) whereas
we and others have found that repeated thawing
significantly affects the results (unpublished results).
Average intraassay variation was <5% and
interassay variationwas 8%. OneU/L approximates
to 0.1 mg/dl of Lp(a) protein.
Concentration of apo-A-1 was determined with a
competitive radioimmunoassay using spedfic
monoclonal antibodies (Pharmacia Diagnostic
N
B) with an intraassay variation about 5%. Variable
effect of storage on the immunoreactivity of apo–
A-I has been described [7], but our results are
similar to reported values using the same
radioimmunoassay. [8] . Apo-B concentration was
determinedwith two site immunoradiometric assay
using two different monoclonal antibodies in ex–
cess Pharmacia Diagnostic
NB
with intra- and
intraassay coeffident of variance 3 and 4%
respectively.
All partidpants completed a questionnaire
concerning smoking habits and history of coronary
artery disease in first degree relatives. Height and
weight measurements were obtained and body mass
index calculated as kg/m
2
•
Blood pressure was
measuredwith aMercury sphygmomanometer after
a five minutes rest.
Causes of death were obtained by reviewing all
death certificates from the start of study until April
1990. In addition, all autopsy reports available
(59%) were reviewed. Allepisodes of acute myo–
cardial infarction were registered for the whole
country by the leelandie MONICA study group.
Cox's proportional hazards mode! was used to
calculate beta-coefficients for predieting risk of
myocardial infarction. [9] .
47
