Doi:10.1016/j.atherosclerosis.2006.10.032

Moderately elevated plant sterol levels are associated with reduced cardiovascular risk—The LASA study Klaus Fassbender , Dieter L¨utjohann , Miranda G. Dik , Marijke Bremmer , Jochem K¨onig , Silke Walter , Yang Liu , Maryse Leti`embre , Klaus von Bergmann , Cees Jonker a Department of Neurology, University Clinic of the Saarland, Kirrberger Str., D-68421 Homburg, Germany b Department of Statistics, University Clinic of the Saarland, Germany c Department of Clinical Pharmacology, University of Bonn, Sigmund-Freud-Str. 25, D-53105 Bonn, Germany d EMGO Institute, VU University Medical Center, Amsterdam, The Netherlands Received 7 July 2006; received in revised form 17 October 2006; accepted 30 October 2006 Abstract
Functional foods with supplementation of plant sterols are already used by millions of people. However, at the same time it is current scientific thinking that elevation of plant sterols in the circulation causes coronary heart disease. Therefore, this study aimed to define the riskfor coronary heart disease associated with moderately high plant sterol plasma levels in a cohort of elderly.
In this study, we evaluated the association between plant sterols and coronary heart disease in a cohort of 1242 subjects older than 65 years, participating at the Longitudinal Aging Study Amsterdam (LASA). Concentrations of sitosterol, campesterol, brassicasterol and stigmasterolwere assessed using highly sensitive and specific gas chromatography–mass spectrometry-selected ion-monitoring. Plant sterol concentrations(and their ratios to cholesterol) were slightly, however, significantly lower in patients with coronary heart disease. Moreover, high plasmaconcentrations of a marker plant sterol, sitosterol, were associated with a markedly reduced risk for coronary heart disease (OR 0.78, CI0.62–0.98, p < 0.05). In contrast neither plant stanols (sitostanol or campestanol) nor the cholesterol synthesis markers (lathosterol, lanosteroland desmosterol) nor their ratios to cholesterol were significantly different in the study groups.
These data suggest that plant sterols could have neutral or even protective effects on development of coronary heart disease, which have to be confirmed in interventional trials.
2006 Elsevier Ireland Ltd. All rights reserved.
Keywords: Plant sterols; Coronary heart disease; Cholesterol; Atherosclerosis 1. Introduction
been shown to reduce the risk of coronary heart disease Recent work suggests that plant sterol esters such as sitosterol High levels of total cholesterol and low-density lipopro- and campesterol esters or stanol (sitostanol and campestanol) tein cholesterol are risk factors for coronary heart disease, esters also act as lipid-lowering agents y competing with which is a major cause of morbidity and mortality in the cholesterol for incorporation into micelles, thereby decreas- Western world. Lipid-lowering agents such as statins have ing the intestinal absorption of cholesterol For thisreason, in an effort to reduce cardiovascular risk through simple dietary intervention, margarines and other foods con- Corresponding author. Tel.: +49 6841 1624102; fax: +49 6841 1624137.
∗∗ Corresponding author. Tel.: +49 228 287 15272/14027; taining plant sterols or plant stanols have been introduced into the market within the last 10 years Currently, the pathophysiological significance of plant sterols is under intense discussion, as there are indications 0021-9150/$ – see front matter 2006 Elsevier Ireland Ltd. All rights reserved.
doi: K. Fassbender et al. / Atherosclerosis 196 (2008) 283–288 that plant sterols, once absorbed, contribute to atheroscle- vascular disease, inspection of medication and information rosis. Evidence in support of this hypothesis comes from obtained from subjects’ general practitioners Coro- experimental studies showing that plant sterols reduce the nary heart disease was considered present when a diagnosis life-span of stroke-prone spontaneously hypertensive rats of angina pectoris and/or a history of myocardial infarction although there is also evidence that this observations was confirmed. Angina pectoris was defined by the presence are related to factors other than plant sterols , of the following: (1) self-reported cardiac disease with symp- there are observations that patients with mutations in ABCG5 toms of pain or a heavy, uncomfortable feeling in the chest or ABCG8 associated with sitosterolemia suffer of premature during exertion that disappeared within 10 min of stopping or coronary artery disease Moreover, increased plasma taking sublingual nitroglycerine; (2) the current use of nitro- plant sterol levels have been associated with higher inci- glycerine; (3) a confirmed cardiac disease diagnosis given dence of coronary heart disease with a positive by the subject’s general practitioner. A history of myocar- family history for such condition with deposition dial infarction was considered present when (1) the subject of those sterols in atheromatous tissues Another reported having had a myocardial infarction in the past and study, however, suggested neutral effects of sitosterolemia on (2) the general practitioner reported an earlier diagnosis of atherosclerosis or a family history for coronary heart disease myocardial infarction Moreover, peripheral artery dis- ease was diagnosed with self-reported pain with walking and Together, many earlier studies suggest a detrimental effect with disappearance when standing still. Diagnosis of cere- of elevated plant sterol levels, arguing against a use of plant brovascular disease (stroke or transient ischemic attack) was sterol-enriched margarines that are currently consumed by made based on the medical history provided by the subjects millions of people. In this study, we examined the risk for combined with information obtained from the subject’s gen- coronary heart disease associated with moderately increased eral practitioner An algorithm for diabetes mellitus, plant sterol or stanol levels in a cohort of elderly subjects including subjects information, information of the general participating in a large community-based study, that were practitioner and use of antidiabetics was made Sub- na¨ıve for both, statins and sterol/stanol-containing functional jects were classified as either normotensive (<130/85 mmHg) or hypertensive (≥130/85 mmHg). Blood pressure wasmeasured in a sitting position using a standard mercurysphygmomanometer.
2. Methods
The plasma samples were kept frozen between 1995/1996 and the date of analysis at −80 ◦C. Prior to the work- up procedure, butylated hydroxytoluene was added asantioxidant. Cholesterol levels were determined by highly The study population of this cross-sectional study con- sensitive gas chromatography-flame ionization detection sisted of 1242 subjects, aged between 65 and 89 (median and not by enzymatical methods, which cannot differ 75) years, participating at the Longitudinal Aging Study between cholesterol and side-chain substituted choles- Amsterdam (LASA) in which blood samples were col- terols, like campesterol (24-methylcholesterol) and sitosterol lected in 1995/1996. The LASA study is an interdisciplinary (24-ethylcholesterol). Plasma concentrations of the plant cohort study on predictors and consequences of changes sterols sitosterol, campesterol, brassicasterol and stigmas- in autonomy and well-being in the aging population in terol, the plant stanols, campestanol and sitostanol and the The Netherlands A random sample of community- cholesterol precursors lathosterol, lanosterol and desmos- dwelling older persons was drawn from three regions in The terol were assessed using an ultra-sensitive and highly Netherlands, representative of the Dutch population. This specific gas chromatography–mass spectrometry selective population-based sample also included institutionalized sub- ion-monitoring method as previously described jects. Respondents with any use of cholesterol lowering drugs of all sterols was confirmed by comparison with the full- (n = 50) were excluded. At this time, only few Dutch citizens scan mass spectra of the authentic compounds. The intra- were on such drugs and plant sterol/stanol-enriched foods still and inter-assay CV’s for all sterols was below 3%. Accu- were not yet introduced onto the market at all. In 5 subjects, racy of the method was established by recovery experiments, BMI values were missing, and in 11 subjects blood pressure day to day variations (CV below 3%), limit of detection and values were absent. Written informed consent was obtained limit of quantification far below of the present concentra- from all respondents. The study was approved by the Medi- tions (<0.001 mg/dL for each sterol) Storage stability cal Ethics Committee of the Vrije University Medical Center of sterols like cholesterol, plant sterols and stanols as well as cholesterol precursors during long-term storage (freezing)was evaluated from repeated thawing and analysis after 1, 5 and 10 years from a set of samples which where takenon 3 different days within 1 week in 1995 in one patient Diagnoses of cardiovascular diseases were ascertained by with phytosterolemia, her heterozygous parents and her sis- combining three data resources: self-reported symptoms of ter Absolute concentrations of cholesterol, plant sterol K. Fassbender et al. / Atherosclerosis 196 (2008) 283–288 and stanols, and cholesterol precurosrs were decreased by sitosterol concentrations would be associated with a 22% maximum 2% after 5 years and 2.6% after 10 years.
reduction of the risk for coronary heart disease.
3.2. Association between plant sterol levels and further manifestations of vascular disease and epidemiologicalvariables For non-parametric comparison of groups, we used the Mann–Whitney U-test with a Bonferroni correction. For Sitosterol concentrations were higher in females than correlations, the Spearman’s correlation coefficient was cal- in males (0.33 ± 0.17 ␮g/dL versus 0.32 ± 0.17 ␮g/dL, culated. Logistic-regression analysis was used to estimate p < 0.001) and lower in diabetic than in non-diabetic subjects odds ratios (OR) of plant sterols and conventional vascular (0.28 ± 0.16 ␮g/dL versus 0.33 ± 0.17 ␮g/dL, p < 0.001). In risk factors modeling the variables, sitosterol, age (males), contrast, sitosterol concentrations were altered neither in age females), sex, cholesterol, D. mellitus, smoking and hypertensive subjects nor in smokers.
hypertension. Sitosterol and cholesterol concentrations were Moreover, 58 patients had peripheral artery disease and logarithmized to the base 10 and divided by log(2) such 96 suffered of cerebrovascular disease. Sitosterol concen- that odds ratios confer the change in risk associated with trations were decreased in subjects with peripheral artery disease (p < 0.05) but not in those with cerebrovascular dis-ease (p = 0.239).
3. Results
3.3. Association between stanols and markers ofcholesterol synthesis and coronary heart disease 3.1. Association between plant sterol concentrations Concentrations of the plant stanols, sitostanol and campestanol (including their ratios to cholesterol) did notsignificantly differ in the study groups Moreover, shows baseline demographic data. Of the 1192 cholesterol precursors, surrogate the markers of cholesterol subjects, 125 patients had coronary heart disease. Plasma synthesis, lathosterol, lanosterol and desmosterol did not sig- concentrations of all plant sterols studied were significantly nificantly differ in patients with or without coronary heart lower in those patients compared to subjects without coronary disease (The ratio of lathosterol to campesterol, heart disease Correcting for cholesterol by calcu- indicating cholesterol synthesis versus absorption showed a lation of a ratio of the sterol to cholesterol levels revealed the significant lower value in the group of coronary heart dis- same highly significant differences. Plasma levels of plant ease. However, this finding is mainly reasoned by the slightly sterols closely correlated with each other, e.g., sitosterol lev- higher levels in the plant sterol marker, campesterol.
els strongly correlated with those of campesterol (r = 0.94, Weak inverse correlations between the concentrations of p < 0.001), brassicasterol (r = 0.75, p < 0.001) and stigmas- lathosterol and those of sitosterol (r = −0.22, p < 0.001), terol (r = 0.80, p < 0.001). On this basis and because of its campesterol (r = −0.22, p < 0.001), stigmasterol (r = −0.17, currently discussed clinical relevance we selected sitosterol p < 0.001) and brassicasterol (r = −0.09, p < 0.005) were as marker plant sterol for further logistic regression analysis.
Logistic regression analysis modeling the variables sitos- terol, age, sex, hypertension, cholesterol, diabetes andsmoking clearly showed that higher plasma levels of sitosterol 4. Discussion
are associated with a significantly reduced risk for coronaryheart disease (OR 0.78, CI 0.62–0.98, p = 0.030) ( Enrichment of the diet with plant sterols lowers absorp- Calculating the OR, we observed that two-fold increase of tion of cholesterol in the intestine owever, the clinical Table 1Baseline characteristics in the study populations (n = 1192) K. Fassbender et al. / Atherosclerosis 196 (2008) 283–288 Table 2Concentrations of cholesterol, plant sterols and stanols and cholesterol precursors in the study populations (mean ± S.D.) a “R Sterol (Stanol)” indicates ratio of the specific sterol or stanol to cholesterol.
* p < 0.005, compared to values in subjects without coronary heart disease.
** p < 0.001, compared to values in subjects without coronary heart disease.
effect of the associated increase in blood concentrations of but – in contrary – with a reduced risk for coronary heart dis- plant sterols is a matter of much current controversy, as ease. Calculating the OR, we observed that two-fold increase patients with genetically determined sitosterolemia exhibit of sitosterol concentrations would be associated with a 22% reduction of the risk for coronary heart disease. In contrast, The results of this community-based study on elderly sub- neither stanols nor markers of cholesterol synthesis were sig- jects show that plant sterol concentrations (and their ratios to nificantly different in subjects with or without coronary heart cholesterol) were significantly increased in patients without disease, arguing against major detrimental effects of stanols.
coronary heart disease. Moreover, high blood concentrations Thus, low concentrations of plant sterols may represent of plant sterols were not associated with an increased risk, a novel independent vascular risk factor. The discrepancy Table 3Modulation of the risk for coronary heart disease by sitosterol and conventional risk factors a Logistic-regression analysis was used to estimate odds ratios (OR) of plant sterols and conventional vascular risk factors modeling the variables, sitosterol, age (males), age females), sex, cholesterol, diabetes mellitus, smoking and hypertension.
b Cholesterol and sitosterol concentrations were logarithmized to the base 10 and divided by log(2) such that OR confer the change in risk associated with K. Fassbender et al. / Atherosclerosis 196 (2008) 283–288 between the protective effects of plant sterols shown here that those effects are unrelated to cholesterol lowering but a and the proartherogenic effects of sitosterolemia due to consequence of further beneficial actions of those drugs.
mutations in ABCG5 or ABCG8 be explained This demonstration of an association between moderately by a dose-relationship of plant sterols, being beneficial in increased plant sterol plasma concentrations and low risk for the moderate range, but detrimental in the extremely high coronary heart disease may contributes to the understand- range (50–200-fold increase in genetically determined sitos- ing of the significance of dietary plant sterols for the human terolemia Alternatively, additional abnormalities organism, although the effects of plant sterol supplementation associated with those mutations may explain the premature on development of cardiovascular disease can only be verified coronary heart disease in sitosterolemia. The different results by further mechanistic experimental work and prospective obtained in this study may – in part – also be explained by use of various methodologies. Here, we used combined gaschromatography–mass spectrometry-selected ion monitoringin order to reach the highest specificity for quantification Acknowledgements
of sterols, excluding any co-eluting compounds like choles-terol precursors or other plant sterols appearing at the same This work was supported by the Dutch Ministry of Health, retention time. Lanosterol is usually poorly separated and Welfare and Sports, the Vrije University Amsterdam, The in part hidden under the peak of sitosterol using common Netherlands and the “Hans und Ilse Breuer-Stiftung”. We are GC columns. Using mass spectrometry-selected ion mon- grateful to Anja Kerksiek and Silvia Friedrichs for technical itoring detection, the compounds are finally separated by characteristic mass unit fragments, which is not possible withflame-ionization detection methods used in previous studies.
As major advantage, the study samples were obtained at References
1995/1996, at a time were statins were very rarely used andplant sterol supplemented functional foods were still not on [1] Heart Protection Study Collaborative Group. MRV/BHF heart pro- the market. Later studies (and also follow-up of this study) tection study of cholesterol lowering with simvastatin in 20,536 would meet considerable difficulties in controlling for such high-risk individuals: a randomized placebo-controlled trial. Lancet treatments and, i.e., consumption of functional foods that are most common now. Thus, individual plant sterol levels in this [2] Law M. Plant sterol and stanol margarines and health. BMJ study are only determined by the genetic background and by [3] Cater NB, Garcia-Garcia AB, Vega GL, Grundy SM. Responsiveness the normal diet. A limitation of this study is the lack of explicit of plasma lipids and lipoproteins to plant stanol esters. Am J Cardiol information of intake of vegetable food at this time, impair- ing to differentiate effects of plant sterols and of potentially [4] Miettinen TA, Puska P, Gylling H, Vanhanen H, Vartiainen E. Reduc- associated vegetarian life style. However, recent work sup- tion of serum cholesterol with sitostanol-ester margarine in a mildlyhypercholesterolemic population. N Engl J Med 1995;333:1308– ports the view that even vegetarian life style has no major effects on plant sterol levels and that the genetical predispo- [5] Huang MZ, Watanabe S, Kobayashi T, et al. Unusual effects of some sition is the major determinant of high plant sterol levels in vegetable oils on the survival time of stroke-prone spontaneously hyper- absence of a use of highly plant sterol-enriched functional tensive rats. Lipids 1997;32:747–51.
[6] Ikeda I, Nakagiri H, Sugano M, et al. Mechanisms of phytosterolemia in stroke-prone spontaneously hypertensive and WKY rats. Metabolism Here, the association between cholesterol and coronary heart disease was lower than expected in light of previous [7] Tatematsu K, Fuma SY, Nagase T, Ichikawa Y, Fujii Y, Okuyama H.
studies, in which a 1% reduction of cholesterol concentrations Factors other than phytosterols in some vegetable oils affect the survival corresponded approximately to a 3% risk decrease of SHRSP rats. Food Chem Toxicol 2004;42:1442–51.
conflicting result may be explained by the focus of this study [8] Shulman RS, Bhattacharyya AK, Connor WE, Fredricksen DS. Beta- sitosterolemia and xantomatosis. N Engl J Med 1976;294:482–3.
on elderly subjects (aged 65–90 years), a study population in [9] Lee MH, Lu H, Hazard S, et al. Identification of a gene, ABCG5, which the role of cholesterol as a vascular risk factor is known important in the regulation of the dietary cholesterol absorption. Nat to be much more complex Indeed, in studies focussing on aged subjects (late 1970s and older), manifestations of [10] Bj¨orkhem I, Boberg K, Leitersdorf E. Inborn errors in bile acid biosyn- coronary heart disease have been shown to be poorly or not thesis and storage of sterols other than cholesterol. In: Scriver C,Beaudet A, Sly W, Valle D, editors. The metabolic and molecular basis of inherited disease. New York: McGraw-Hill; 2001.
Concentrations of sitosterol were decreased also in [11] Glueck CJ, Speirs J, Tracy T, et al. Relationship of serum plant sterols peripheral artery disase but not in cerebrovascular dis- (phytosterols) and cholesterol in 595 hypercholesterolemic subjects, eases. This corroborates the reported very weak or absent and familial aggregation of phytosterols, cholesterol, and premature relationship between lipids and stroke a highly hetero- coronary heart disease in hyperphytosterolemic probands and their first-degree relatives. Metabolism 1991;40:842–8.
geneous disease entity. Although statins (in contrast to other [12] Sutherland WHF, Williams MJA, Hye ER, et al. Association of plasma cholesterol-lowering drugs) were effective in prevention of noncholesterol sterol levels with severity of coronary artery disease.
stroke in the Heart Protection Study it has been shown Nutr Metab Cardiovasc Dis 1998;8:386–91.
K. Fassbender et al. / Atherosclerosis 196 (2008) 283–288 [13] Rajaratnam RA, Gylling H, Miettinen TA. Independent association of [22] L¨utjohann D, Bj¨orkhem I, Ose L. Phytosterolemia in a Norwegian fam- serum squalene and noncholesterol sterols with coronary artery disease ily: diagnosis and characterization of the first scandinavian case. Scand in postmenopausal women. J Am Coll Cardiol 2000;35:1185–91.
[14] Assmann G, Cullen P, Erbey J, et al. Elevation in plasma sitosterol [23] Clarenbach JL, Reber M, L¨utjohann D, von Bergmann K, Sudhop T.
concentration is associated with an increased risk for coronary events The lipid-lowering effect of ezetimibe in pure vegetarians. J Lipid Res in the PROCAM study. Circulation 2003;108(Suppl.):3300.
2006;(September) [Epub ahead of print].
[15] Sudhop T, Gottwald BM, von Bergmann K. Serum plant sterols [24] Berge KE, von Bergmann K, L¨utjohann D, et al. Heritability of plasma as a potential risk factor for coronary heart disease. Metabolism noncholesterol sterol and relationship to DNA sequence polymorphism in ABCG5 and ABCG8. J Lipid Res 2002;43:486–94.
[16] Mellies MJ, Ishikawa TT, Glueck CJ, et al. Phytosterols in aortic tissue [25] National Research Council, Committee on Diet and Health: Diet and in adults and infants. J Lab Clin Med 1976;88:914–21.
Health. Implications for reducing chronic disease risk. Washington, [17] Miettinen TA, Railo M, Lepantalo M, Gylling H. Plant sterols in serum and in atherosclerotic plaques of patients undergoing carotid [26] Garber AM, Browner WS, Hulley SB. Clinical screening in asymptotic endarterectomy. J Am Coll Cardiol 2005;45:1794–801.
adults, revisited. Ann Int Med 1996;124:518–31.
[18] Wilund KR, Yu L, Xu F, et al. No association between plasma levels of [27] Hulley SB, Newman TB. Cholesterol and mortality. 30 years of plant sterols and atherosclerosis in mice and men. Arterioscler Thromb follow-up from the Framingham study. JAMA 1987;257:2176– [19] Smit JH, de Vries MZ, Poppelaars JL. Data collection and fieldwork [28] Kronmal RA, Cain KC, Ye Z, Omenn GS. Total serum cholesterol procedures. In: Deeg DJ, Beekman AT, Kriegsman DM, Westendorp-de levels and mortality risk as a function of age. A report based on the Sieriere M, editors. Autonomy and well-being in the aging population Framingham data. Arch Intern Med 1993;153:1065–73.
II: report from the Longitudinal Aging Study Amsterdam 1992–1996.
[29] Krumholz HM, Seeman TE, Merrill SS, et al. Lack of association Amsterdam, The Netherlands: VU University Press; 1998. p. 9–20.
between cholesterol and coronary heart disease mortality and mor- [20] Penninx BW, Beekman AT, Honig A, et al. Depression and cardiac bidity and all-cause mortality in persons older than 70 years. JAMA mortality: results from a community-based longitudinal study. Arch [30] Prospective Studies Collaboration. Cholesterol, diastolic blood pres- [21] Sudhop T, L¨utjohann D, Kodal A, et al. Inhibition of intestinal choles- sure, and stroke: 13,000 in 450,000 people in 45 prospective cohorts.
terol absorption by ezetimibe in humans. Circulation 2002;106:1943–8.

Source: http://www.nufs.sjsu.edu/clariebh/sterols.pdf