The main finding of this study was that higher level of pessimism predicted the incidence of CHD after adjusting to other predictors of CHD. Significant proportion of this effect was
mediated by increased inflammation among those with more pessimism.
We are reporting for the first time the possible mediator effect of hs-CRP in the context of
pessimism and cardiovascular risk. Pessimism was an independent predictor of higher
incidence of CHD in both models applied. This finding is in line with a large body of evidence
connecting pessimism and higher incidence of CHD , , , , . Hs-CRP, a marker
of inflammation, and a known predictor of cardiovascular disease, was also connected to
incidence of CHD in the present population. According to our analysis hs-CRP mediated nearly
30% of the pessimism related effect on CHD.
Positive affective states have been connected with lower hs-CRP levels  while increasing
severity of depression seems to correlate with an elevated hs-CRP level , . In a
research derived from the Heart and Soul Study, the connections between depression and
inflammation, between depression and CHD, and between inflammation and CHD were
affirmed. Contrary to our positive findings about the inflammation mediating the effect of
psychosocial risk factor on the incidence of CHD, investigators of Heart and Soul Study found
no evidence of inflammation operating as a mediator between depression and CHD .
Parallel findings have also been made about optimism. A study about the optimistic
perception of one’s CVD risk found lower rates of adverse effects of CVD among those with
more optimism but this connection could not be explained by the mediating effect of the level
of inflammation . In that study, optimism was measured by asking the test subjects what
they think about their risk of developing CHD, so the question evaluated self-rated health
rather than optimism.
Even if both LOT and LOT-R were thought to be unidimensional scales (i.e. the result of the test was given as a combination of the answers to the pessimistically and optimistically
oriented statements), later studies have suggested that they may have two separate
independent dimensions, namely optimism and pessimism. Dividing the optimism scaling
from pessimism and examining the results separately seems to lead to better prediction of
outcomes , . In a bipolar univariate model, optimism and pessimism may hide some of
each other’s results. It has also been suggested that dispositional optimism could be a
unidimensional continuum, but LOT-R divides it into two independent subscales, optimism
and pessimism  and the questions worded negatively might be better at measuring this
personal trait than the optimistically oriented questions. According to this theory the
questions worded positively may not give consistent results, leading to the weak statistical
power of the optimism subscale .
In our earlier study we demonstrated that in this study group, optimism and pessimism seem
to be separate variables , a finding that has also been made in many other study groups
, , . In the present study, the two-factor model of optimism and pessimism – i.e.
optimism and pessimism being two separate and independent variables instead of one bipolar
factor – was confirmed. While the connection between pessimism and CHD appeared to be
solid, optimism did not associate with the incidence of CHD.
Optimism and pessimism characterize an individual in ordinary situations. Like other
personality traits, their development appears to be influenced by both heritage and
environment , . Unlike mood, for example, they seem to be stable once they have
evolved, and they remain the same in different situations and over time, regardless of the
negative or positive incidents , .
The association between psychosocial factors and CHD seems to be quite confident, and our study endorses the need to continue the research in this area. Our hypothesis of inflammation mediating the connection between pessimism and CHD was confirmed to a marked degree. Finding the other mediating factors between pessimism and CHD could help in understanding CHD, its aetiology and mechanism, which in turn could help in preventing and treating CHD.
The strength of our study is in its design. The study group can be seen as comprehensive one.
Ten years is a relatively long time, and the prospective nature of the study makes the results
more reliable. The use of a well-known test pattern (LOT-R) in determining optimism and
pessimism makes our study more convincing. Separating optimism from pessimism seemed
to clarify our results. CRP has a long plasma half-life and relative stability as a frozen sample,
which makes it quite easy and reliable to measure .
There are a few shortcomings in this study. One of them is that we could use only self-reports
for some variables, which can lead to biased information. Self-reports were used in the
variables concerning life habits (e.g. smoking and the use of alcohol). However, it has been
demonstrated that alcohol consumption can be estimated quite accurately using self-reports
, , ,  and while self-reports concerning smoking status are usually not so
reliable, they are very useful when the respondents are elderly , like the participants in
our study. We had access to the Records of Causes of Deaths, and thus we could determine the
number of people who died due to CHD, but the prevalence of CHD at the beginning of the
study and the number of new cases of CHD during the follow-up among those who were alive
at the end of the study were based on self-reports. However, self-reports are quite reliable like
earlier mentioned, and the numbers concerning CHD received by self-reports are quite similar
to the prevalence and incidence rates of CHD that can be calculated from the official statistics
for the same-aged population in Finland , .
It is also expected that there were some people with new CHD among those who died during
the follow-up due to some other reason than CHD, and those individuals could not be included
in the study. It also seems that poorly functioning and institutionalized persons had a lower
participation rate than community-dwelling subjects, which can have an effect on the results
. In addition, the level of CRP may have been high for temporary reasons in some cases, which might have lightened our finding. Another possible form of bias is related to the fact that the levels of hs-CRP were analysed in samples that had been frozen for more than 10 years.
Consequently, there is a risk that the absolute levels of hs-CRP could have been affected by
having been taken from frozen – rather than fresh – samples. However, it has been found that
CRP is relatively stable as a frozen sample  and the freezing time was the same for all