Professional Documents
Culture Documents
org
OBSTETRICS
Maternal C-reactive protein and developmental
programming of atherosclerosis
Antonio Liguori, MD; Francesco P. D’Armiento, MD; Antonio Palagiano, MD; Wulf Palinski, MD; Claudio Napoli, MD, PhD
OBJECTIVE: Maternal hypercholesterolemia during pregnancy en- RESULTS: CRP level was elevated in hypercholesterolemic mothers
hances the susceptibility to atherosclerosis in their offspring by and showed significant correlation with atherogenesis in children in
oxidation-dependent mechanisms. The present study investigated univariate and multivariate analysis. However, many hypercholester-
whether maternal C-reactive protein (CRP) level, which is an indi- olemic mothers did not have elevated CRP levels. Smoking only cor-
cator of inflammation and cardiovascular risk, or smoking, which related in univariate analysis.
enhances oxidative stress, predict the in utero programming of
atherosclerosis. CONCLUSION: CRP level during pregnancy is a predictor of increased
atherogenesis in children of hypercholesterolemic mothers, albeit a weaker
STUDY DESIGN: Subsets of patients from the Fate of Early Lesions in one than maternal cholesterol. In the presence of hypercholesterolemia,
Childhood study (156 normocholesterolemic children) were examined maternal smoking does not further enhance atherogenic programming.
at autopsy, classified by maternal cholesterol levels during pregnancy.
Maternal CRP level was correlated with maternal cholesterol and aortic Key words: atherosclerosis, CRP, developmental programming,
atherosclerosis of children. inflammation, maternal hypercholesterolemia, oxidative stress
Cite this article as: Liguori A, D’Armiento FP, Palagiano A, Palinski W, Napoli C. Maternal C-reactive protein and developmental programming of
atherosclerosis. Am J Obstet Gynecol 2008;198:281.e1-281.e5.
particular, multiple regression analysis 0.14 mg/dL; P ⬍ .000001; Figure 1). In-
FIGURE 2
of the correlation between maternal CRP terestingly, less than one-half of the hy-
Maternal plasma cholesterol
level, smoking, and total cholesterol with percholesterolemic mothers had sub-
level correlates with maternal
offspring atherosclerosis), data from stantially elevated CRP levels (⬎ 8
CRP level during pregnancy
both groups were combined and ana- mg/dL).
lyzed together, and actual maternal cho- The correlation between maternal
lesterol levels were used. Case numbers CRP and plasma cholesterol levels in a
for each analysis vary, because not all subset of mothers (n ⫽ 89) for which
maternal records contained all of the pa- both values could be obtained from
rameters that were assessed. medical records of the same prenatal ex-
amination is shown in Figure 2. The per-
centage of mothers who smoked during
Statistical analysis
pregnancy was much greater in the hy-
Maternal total cholesterol and CRP lev-
percholesterolemic group than in the
els were compared by unpaired t test and
normocholesterolemic group (58.6% vs
linear correlation. The cumulative lesion
38.5%), whereas the average number of
Data represent all mothers for whom both data areas in the aortic arch and abdominal
cigarettes per day per smoker was simi-
were available at the same time point (n ⫽ 89). aorta are correlated highly and were con-
lar. CRP also correlated significantly
In contrast to all other analyses, cholesterol densed into a single parameter, with the
with maternal smoking in univariate
levels that were obtained at the same time point use of principal-components analysis of
analysis (R2 ⫽ 0.223; P ⬍ .0005). When
as CRP level were used instead of mean cho- log-transformed lesion sizes. The first
CRP level was used as a dependable pa-
lesterol levels. principal component accounted for
rameter in multivariate analysis, both
Liguori. Maternal CRP and atherosclerosis. Am J Obstet more than 98.5% of the variation of the
maternal plasma cholesterol level and
Gynecol 2008. data (Eigen value, 1.97). We then per-
smoking remained significant (R2 ⫽
formed multiple stepwise linear regres-
0.408; P ⬍ .0005; standardized beta val-
sion analysis with maternal risk factors,
tively. This assay had a sensitivity of ues 0.468 and 0.370, respectively).
using pooled data from children of both
0.025 mg/L. The FELIC study had indicated a
normocholesterolemic and hypercho-
We initially compared CRP levels be- strong correlation of atherosclerosis
lesterolemic mothers. The criteria to be
tween the 2 maternal groups defined in with age (Figure 3), which must be kept
entered into the stepwise model were an
the Lancet study. For further analyses (in in mind when assessing correlations of
enter value of 0.05 and an elimination
maternal factors with non–age-cor-
value of 0.10. From the resultant model,
FIGURE 3 rected atherosclerosis data, such as the
we tested individual covariates with
Atherosclerosis progresses composite measure of atherosclerosis in
Bonferroni-adjusted criteria for signifi-
much faster in children the arch and abdominal aorta of chil-
cance. P ⬍ .05 was considered signifi-
of mothers who were dren. Nevertheless, in addition to factors
cant. All statistical evaluations were per-
hypercholesterolemic previously established as significant (ie,
formed with SPSS statistical software (v
during pregnancy the age of the child [R2 ⫽ 0.536; P ⬍
13; SPSS Inc, Chicago, IL). Data are re-
.0005], the maternal group [R2 ⫽ 0.368;
ported as means ⫾ SEM.
P ⬍ .0005], and maternal cholesterol
The FELIC study and its present ex-
level [R2 ⫽ 0.343; P ⬍ .0005]), maternal
tension were approved by the human
CRP level (R2 ⫽ 0.173; P ⬍ .0005; Figure
ethics committee of the University of
4) and maternal smoking (R2 ⫽ 0.084; P
Naples and the University of California,
⬍ .001) showed significant correlation
San Diego. Informed consent was ob-
with offspring atherosclerosis in univar-
tained from all participating mothers.
iate analysis.
Significant parameters (child age, ma-
R ESULTS ternal total cholesterol, CRP, and smok-
By definition, maternal cholesterol levels ing) were then included in a stepwise
in the second to third trimester of preg- multiple regression analysis. Note that,
Examples shown are the cumulative areas of
nancy were significantly greater in the in contrast to the FELIC study, we used
atherosclerotic lesions in 30 cross-sections
hypercholesterolemic than the normo- the single maternal cholesterol value that
through the aortic arch (n ⫽ 97 and 59 for
cholesterolemic maternal group (194.9 was measured at the same time as CRP
children of hyper- and normocholesterolemic
⫾ 2.9 mg/dL vs 279.9 ⫾ 3.4 mg/dL; P ⬍ level, instead of the (highly correlated)
mothers, respectively).8
Liguori. Maternal CRP and atherosclerosis. Am J Obstet
.000001). Maternal CRP levels were also maternal group, which was defined by
Gynecol 2008. much greater in hypercholesterolemic the average maternal cholesterol level
mothers (6.93 ⫾ 0.36 mg/dL vs 4.68 ⫾ during multiple prenatal examinations.
TABLE
Stepwise linear regression analysis with the use of a composite measure of atherosclerosis
in the aortic arch and abdominal aorta as outcome parameter and factors
that were correlated significantly with it in univariate analysis as predictors
Standard error of Significance of Standardized Significance
Model R2 the estimate modela coefficient (beta) of covariatea
1. Age of child 0.489 0.688 .0005 0.699 .0005
................................................................................................................................................................................................................................................................................................................................................................................
2. Age of child 0.872 0.343 .0005 0.732 .0005
................................................................................................................................................................................................................................................................................................................................................................................
Maternal cholesterol 0.622 .0005
................................................................................................................................................................................................................................................................................................................................................................................
3. Age of child 0.884 0.331 .0005 0.740 .0005
................................................................................................................................................................................................................................................................................................................................................................................
Maternal cholesterol 0.562 .0005
................................................................................................................................................................................................................................................................................................................................................................................
Maternal CRP 0.116 .009
................................................................................................................................................................................................................................................................................................................................................................................
Maternal cholesterol level, age of child, and atherosclerosis, 156 patients; maternal smoking, 154 patients; maternal CRP level, 136 patients.
a
Probability value.
Liguori. Maternal CRP and atherosclerosis. Am J Obstet Gynecol 2008.
Limitations include the retrospective 5. Napoli C, D’Armiento FP, Mancini FP, et al. 19. Wolbink GJ, Brouwer MC, Buysmann S, ten
nature of the FELIC study and the ensu- Fatty streak formation occurs in human fetal aor- Berge IJ, Hack CE. CRP-mediated activation of
tas and is greatly enhanced by maternal hyper- complement in vivo: assessment by measuring
ing need to rely on medical records for cholesterolemia. Intimal accumulation of low den- circulating complement–C-reactive protein
maternal data during pregnancy. The sity lipoprotein and its oxidation precede complexes. J Immunol 1996;157:473-9.
fact that these data were generated by a monocyte recruitment into early atherosclerotic 20. Li SH, Szmitko PE, Weisel RD, et al. C-re-
number of different clinical laboratories lesions. J Clin Invest 1997;100:2680-90. active protein upregulates complement-inhibi-
over a 14-year period may have increased 6. Napoli C, Witztum JL, Calara F, de Nigris F, tory factors in endothelial cells. Circulation
Palinski W. Maternal hypercholesterolemia en- 2004;109:833-6.
variability. The retrospective nature of hances atherogenesis in normocholesterolemic 21. Pasceri V, Willerson JT, Yeh ET. Direct
the study also made it impossible to sup- rabbits, which is inhibited by antioxidant or lipid- proinflammatory effect of C-reactive protein on
plement maternal data during preg- lowering intervention during pregnancy: an ex- human endothelial cells. Circulation 2000;102:
nancy by measurements of other li- perimental model of atherogenic mechanisms 2165-8.
poproteins, in particular high-density in human fetuses. Circ Res 2000;87:946-52. 22. Wang CH, Li SH, Weisel RD, et al. C-reac-
7. Palinski W, D’Armiento FP, Witztum JL, et al. tive protein upregulates angiotensin type 1 re-
lipoprotein, parameters of oxidative Maternal hypercholesterolemia and treatment ceptors in vascular smooth muscle. Circulation
stress (to validate smoking history), and during pregnancy influence the long-term pro- 2003;107:1783-90.
other parameters of inflammation, such gression of atherosclerosis in offspring of rab- 23. Blaschke F, Bruemmer D, Yin F, et al. C-re-
as isoprostanes or antibody titers to oxi- bits. Circ Res 2001;89:991-6. active protein induces apoptosis in human cor-
dation-specific epitopes. In particular, 8. Napoli C, Glass CK, Witztum JL, Deutsch R, onary vascular smooth muscle cells. Circulation
D’Armiento FP, Palinski W. Influence of mater- 2004;110:579-87.
uterine infections or general infections nal hypercholesterolaemia during pregnancy on 24. Napoli C, Witztum JL, de Nigris F, Palumbo
and other conditions that raise CRP level progression of early atherosclerotic lesions in G, D’Armiento FP, Palinski W. Intracranial arter-
may not have been documented in childhood: Fate of Early Lesions in Children (FE- ies of human fetuses are more resistant to hy-
maternal medical records, and their con- LIC) study. Lancet 1999;354:1234-41. percholesterolemia-induced fatty streak forma-
founding effect may have been underes- 9. Libby P. Inflammation and cardiovascular tion than extracranial arteries. Circulation
disease mechanisms. Am J Clin Nutr 2006; 1999;99:2003-10.
timated. Furthermore, not all parame- 83:456S-60S. 25. Palinski W, Napoli C. Unraveling pleiotropic
ters were available in all of the 156 10. Yamashita T, Freigang S, Eberle C, et al. effects of statins on plaque rupture. Arterioscler
mothers and their children, which re- Maternal immunization programs postnatal im- Thromb Vasc Biol 2002;22:1745-50.
duced the power of the analysis. Finally, mune responses and reduces atherosclerosis 26. Taylor PD, Khan IY, Hanson MA, Poston L.
the incidence of other maternal condi- in offspring. Circ Res 2006;99:E51-E64. Impaired EDHF-mediated vasodilatation in
11. Haverkate F, Thompson SG, Pyke SD, Gal- adult offspring of rats exposed to a fat-rich diet
tions that potentially affect oxidative limore JR, Pepys MB. Production of C-reactive in pregnancy. J Physiol 2004;558:943-51.
stress, such as maternal diabetes melli- protein and risk of coronary events in stable and 27. Khan IY, Dekou V, Douglas G, et al. A high-
tus, was too low to yield reliable data for unstable angina: European Concerted Action fat diet during rat pregnancy or suckling induces
CRP. Ideally, the association of maternal on Thrombosis and Disabilities Angina Pectoris cardiovascular dysfunction in adult offspring.
risk factors with atherogenic program- Study group. Lancet 1997;349:462-6. Am J Physiol Regul Integr Comp Physiol
12. Ridker PM, Hennekens CH, Buring JE, Rifai 2005;288:R127-33.
ming and coronary heart disease should N. C-reactive protein and other markers of inflam- 28. Bottiglioni F, Guerresi E, Tirelli R, Giordano
be assessed in prospective studies, but mation in the prediction of cardiovascular disease G. On hypercholesteremia of pregnancy. Con-
such studies would have to be very large in women. N Engl J Med 2000;342:836-43. tribution to the study of the phenomenon and its
and span at least 4 decades for cardiovas- 13. Ridker PM, Rifai N, Rose L, Buring JE, Cook current interpretive possibilities. Riv Ital Ginecol
cular outcomes (lesser periods, if surro- NR. Comparison of C-reactive protein and low- 1967;51:20-50.
density lipoprotein cholesterol levels in the pre- 29. Napoli C, de Nigris F, Welch JS, et al. Ma-
gate measures of atherosclerosis are diction of first cardiovascular events. N Engl ternal hypercholesterolemia during pregnancy
considered). f J Med 2002;347:1557-65. promotes early atherogenesis in LDL receptor-
14. Ridker PM, Buring JE, Cook NR, Rifai N. deficient mice and alters aortic gene expression
C-reactive protein, the metabolic syndrome, determined by microarray. Circulation 2002;
REFERENCES and risk of incident cardiovascular events: an 105:1360-7.
1. Barker DJ, Winter PD, Osmond C, Margetts B, 8-year follow-up of 14,719 initially healthy 30. Heitzer T, Ylä-Herttuala S, Luoma J, et al.
Simmonds SJ. Weight in infancy and death from American women. Circulation 2003;107:391-7. Cigarette smoking potentiates endothelial dys-
ischaemic heart disease. Lancet 1989;2:577-80. 15. Libby P, Ridker PM. Inflammation and athero- function of forearm resistance vessels in pa-
2. Barker DJ, Gluckman PD, Godfrey KM, Har- sclerosis: role of C-reactive protein in risk assess- tients with hypercholesterolemia: role of oxi-
ding JE, Owens JA, Robinson JS. Fetal nutrition ment. Am J Med 2004;116(suppl):9S-16S. dized LDL. Circulation 1996;93:1346-53.
and cardiovascular disease in adult life. Lancet 16. Ridker PM, Rifai N, Lowenthal SP. Rapid 31. Yang Z, Knight CA, Mamerow MM, et al.
1993;341:938-41. reduction in C-reactive protein with cerivastatin Prenatal environmental tobacco smoke expo-
3. Palinski W, Napoli C. The fetal origins of ath- among 785 patients with primary hypercholes- sure promotes adult atherogenesis and mito-
erosclerosis: maternal hypercholesterolemia, terolemia. Circulation 2001;103:1191-3. chondrial damage in apolipoprotein E-/- mice
and cholesterol-lowering or antioxidant treat- 17. Engler MM, Engler MB, Malloy MJ, et al. An- fed a chow diet. Circulation 2004;110:3715-20.
ment during pregnancy influence in utero pro- tioxidant vitamins C and E improve endothelial 32. Luo ZC, Fraser WD, Julien P, et al. Tracing
gramming and postnatal susceptibility to function in children with hyperlipidemia: Endothe- the origins of “fetal origins” of adult diseases:
atherogenesis. FASEB J 2002;16:1348-60. lial Assessment of Risk from Lipids in Youth programming by oxidative stress? Med Hypoth-
4. Napoli C, Palinski W. Maternal hypercholester- (EARLY) trial. Circulation 2003;108:1059-63. eses 2006;66:38-44.
olemia during pregnancy influences the later de- 18. Volanakis JE. Complement activation by C- 33. Napoli C, de Nigris F, Palinski W. Multiple
velopment of atherosclerosis: clinical and patho- reactive protein complexes. Ann N Y Acad Sci role of reactive oxygen species in the arterial
genic implications. Eur Heart J 2001;22:4-9. 1982;389:235-50. wall. J Cell Biochem 2001;82:674-82.