effects of consumption of probiotics and prebiotics on serum lipid levels in humans
TRANSCRIPT
259
1040-9238/02/$.50© 2002 by CRC Press LLC
Critical Reviews in Biochemistry and Molecular Biology, 37(4):259–281 (2002)
Effects of Consumption of Probiotics andPrebiotics on Serum Lipid Levels inHumans
Dora I. A. Pereira* and Glenn R. Gibson
Food Microbial Sciences Unit, School of Food Biosciences, The Univ. of Reading,England
* Corresponding author. Food Microbial Sciences Unit, School of Food Biosciences, The University of Reading,Earley Gate, Whiteknights Road, Reading RG6 6BZ, England.
Referee: Dr. Marcel Roberfroid, Universite Catholique de Louvain, Faculte de Medicine, Dept.des Sciences Pharmaceutiques, Avenue E. Mournier 73, UCL/BCTC, B-1200 Bruxelles,Belgium
ABSTRACT: The objective of this article is to review existing studies concerning theeffects of probiotics and prebiotics on serum cholesterol concentrations, with particularattention on the possible mechanisms of their action. Although not without exception,results from animal and human studies suggest a moderate cholesterol-lowering action ofdairy products fermented with appropriate strain(s) of lactic acid bacteria and bifidobacteria.Mechanistically, probiotic bacteria ferment food-derived indigestible carbohydrates toproduce short-chain fatty acids in the gut, which can then cause a decrease in the systemiclevels of blood lipids by inhibiting hepatic cholesterol synthesis and/or redistributingcholesterol from plasma to the liver. Furthermore, some bacteria may interfere with choles-terol absorption from the gut by deconjugating bile salts and therefore affecting themetabolism of cholesterol, or by directly assimilating cholesterol.
For prebiotic substances, the majority of studies have been done with thefructooligosaccharides inulin and oligofructose, and although convincing lipid-loweringeffects have been observed in animals, high dose levels had to be used. Reports in humansare few in number. In studies conducted in normal-lipidemic subjects, two reported noeffect of inulin or oligofructose on serum lipids, whereas two others reported a significantreduction in serum triglycerides (19 and 27%, respectively) with more modest changes inserum total and LDL cholesterol. At present, data suggest that in hyperlipidemic subjects,any effects that do occur result primarily in reductions in cholesterol, whereas in normallipidemic subjects, effects on serum triglycerides are the dominant feature.
KEY WORDS: cholesterol, lactobacilli, bifidobacteria, fructooligosaccharides, fermenteddairy products.
Cri
tical
Rev
iew
s in
Bio
chem
istr
y an
d M
olec
ular
Bio
logy
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f C
alga
ry o
n 09
/10/
13Fo
r pe
rson
al u
se o
nly.
260
I. INTRODUCTION
Coronary heart disease (CHD) is one ofthe major causes of death and disability inindustrialized countries. The results fromseveral epidemiological and clinical studiesindicate a positive correlation between el-evated total serum cholesterol levels, mainlyreflecting the LDL-cholesterol fraction, andrisk of emergence of CHD (Lipid ResearchClinics Program, 1984). It has been shownthat a reduction in total plasma cholesterollevels in populations suffering from primaryhypercholesterolemia can lower the inci-dence of coronary thrombosis in this group(Levine et al., 1995). More recent findings,from large-scale epidemiological surveys,provide evidence that elevated fasting trig-lyceride levels are associated with a greaterrisk of CHD, and that this effect is indepen-dent of any association with low levels ofHDL-cholesterol (Hokanson and Austin,1996). A significant number of studies havealso shown that elevated postprandial trig-lyceride concentrations are observed in sub-jects with CHD (Groot et al., 1991) andtherefore can be useful as predictors of CHDrisk (Patsch, 1992).
Currently, there is extensive interest inthe dietary management of serum choles-terol and triglycerides levels. This is largelydriven by the expense of drug therapy, thelarge numbers of individuals affected andunwanted side effects of such treatments.Current dietary strategies for prevention ofCHD implicate adherence to a low-fat/low-saturated fat diet (Taylor and Williams,1998). Although such diets may present aneffective therapy, they are difficult to main-tain on a long-term basis and their efficacydiminishes over time. As such, new ap-proaches toward the identification of otherdietary means of reducing blood cholesterollevels have been evaluated. These includethe use of soluble fibers, soy protein, plant
sterols, probiotic bacteria, and prebiotic com-pounds (Taylor and Williams, 1998).
The focus of this review is the effects ofprobiotics and prebiotics on blood lipid lev-els in humans, with particular attention givento the mechanisms by which such effectsmight be exerted.
II. PROBIOTICS
The first record of the influence of cer-tain dairy products on blood lipids dates backmore than 30 years. Shaper and colleagues(1963) and later Mann (1974) observed thatmen from the tribes of Samburu and Maasaiwarriors in Africa showed a reduced serumcholesterol after consumption of largeamounts of milk fermented with a wild Lac-tobacillus strain. Since then, the potentialhypocholesterolemic effect of fermented milkproducts containing lactobacilli and/orbifidobacteria has been investigated in di-etary studies using humans and animals.
In recent years the European marketfor products claiming to lower blood cho-lesterol, and thereby contributing toward ahealthier heart, has grown substantially(Young, 1998). For example, Danonelaunched Actimel Cholesterol Control® inBelgium, containing the suggested choles-terol-lowering probiotic Lactobacillus aci-dophilus and the branded prebioticACTILIGHT ® (Beghin-Say), while theDutch company Mona introduced a cul-tured dairy-based drink under the Fysiq®
brand. Fysiq® contains the probiotic Lac-tobacillus acidophilus and the brandedbifidogenic dietary fiber, RAFTILINE®
(Orafti). The Danish company MD Foodsintroduced in Denmark in 1993 a yoghurt-style product Gaio. However, controversystill exists as to the beneficial effects ofsome of these, and other products on thelevels of blood lipids in humans.
Cri
tical
Rev
iew
s in
Bio
chem
istr
y an
d M
olec
ular
Bio
logy
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f C
alga
ry o
n 09
/10/
13Fo
r pe
rson
al u
se o
nly.
261
A. Human Dietary InterventionStudies
Harrison and Peat (1975) found thatserum cholesterol levels in bottle-fed babiesdecreased as the number of Lactobacillusacidophilus in their stools increased. Howardand Marks (1977) observed that the con-sumption of 2 L of whole-fat milk and skimmilk reduced serum cholesterol by 5 and15%, respectively, while intake of the sameamount of milk fat as butter increased se-rum cholesterol by 7%. Yoghurt has alsobeen shown to decrease serum cholesterollevels in humans (Hepner et al., 1979), al-though the effect has sometimes been con-sidered to be transient (Rossouw et al., 1981).Several other studies designed to evaluatethe potential reduction of serum cholesterollevels by the consumption of certain cul-tured dairy products have given variabledata and no firm conclusions can be drawn(Taylor and Williams, 1998). In most ofthese studies a decrease in serum choles-terol was only observed in humans duringthe consumption of very high doses of fer-mented dairy products. Other investigatorsusing more ‘normal’ doses of the fermentedmilk product failed to confirm such find-ings. It was suggested that the contradictoryresults obtained could be, at least in part,related to experimental design (Taylor andWilliams, 1998). Some of the factors ad-dressed were lack of statistical power, useof inadequate sample sizes, failure to con-trol nutrient intake and energy expenditureduring the experiments and variations in thebaseline levels of blood lipids. Theintraindividual variation in blood choles-terol over a few months ranged from 5 to15%.
More recent dietary studies using ran-dom double-blind placebo procedures andhigher ranges of human subjects havereached the same conclusions (Table 1).
Some showed the cholesterol-lowering ef-fect of various fermented milk products,while others failed to demonstrate a signifi-cant effect on cholesterol or lipoprotein lev-els by dietary supplementation of these prod-ucts. Differences in experimental design(type and quantity of the fermented milkproduct; age and sex distribution, and start-ing plasma cholesterol levels of the subjectsstudied; and length of study period), how-ever, make direct comparisons difficult.
Andersson et al. (1995) studied the ef-fect of low-fat milk and fermented low-fatmilk, containing strains of Lactococcus lactisand Lactococcus cremoris, on cholesterolabsorption and excretion in ileostomy sub-jects. In this study, patients consumed 1 L/day of either one of the milk products inaddition to their normal diets in a crossoverdesign procedure of 3 weeks with run-inand run-out periods of 2 weeks, each with 1 Lof lemonade. No significant change in se-rum cholesterol was observed after 3 weekson each milk regimen.
It has been reported that recently a fer-mented milk product (Gaio®) that is pro-duced through the action of a bacterial cul-ture containing a strain of Enterococcusfaecium and two strains of Streptococcusthermophilus (CAUSIDO® culture) was ef-fective in reducing plasma cholesterol atrelatively modest levels of intake. The bac-terial strains contained in this product wereisolated from the intestinal flora of inhabit-ants of Abkhasia (Caucasus), a region re-puted for the longevity of its people andwhere fermented milk is a major part of thediet (Agerbaek et al., 1995). In a study per-formed with Danish middle-aged men a re-duction of around 6% of total plasma cho-lesterol and 10% of LDL-cholesterol wasreported after the consumption of 200 ml/day of this fermented milk product over aperiod of 6 weeks (Agerbaek et al., 1995).No change was observed in HDL-choles-terol or plasma triacylglycerol levels. This
Cri
tical
Rev
iew
s in
Bio
chem
istr
y an
d M
olec
ular
Bio
logy
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f C
alga
ry o
n 09
/10/
13Fo
r pe
rson
al u
se o
nly.
262
TABLE 1Details of Recent Human Dietary Studies on the Evaluation of the Cholesterol-LoweringProperties of Fermented Milk Products
Cri
tical
Rev
iew
s in
Bio
chem
istr
y an
d M
olec
ular
Bio
logy
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f C
alga
ry o
n 09
/10/
13Fo
r pe
rson
al u
se o
nly.
263
study was randomised, double-blind, andplacebo-controlled, with a subject group of58 male volunteers aged 44 years old. In afollow-up study the long-term effect of thisproduct on the level of plasma lipoproteinswas evaluated. Richelsen et al. (1996) ob-served a rapid reduction of LDL-cholesterolafter 1 month of consumption of 200 ml/dof the Gaio® product. However, during long-term intake (6 months) the reduction of LDL-cholesterol was similar to that of placebo(chemically fermented) product. The authorsconcluded that although low-fat milk or fer-mented milk products may have somehypocholesterolemic effects, the tested prod-uct was not superior to the placebo milkproduct used in the study because choles-terol concentrations had also fallen in theplacebo group and attributed this to knownseasonal variations in cholesterol levels(Taylor and Williams, 1998). However, theinterpretation of the results was made diffi-cult by the fact that during the second halfof the study concentrations of Enterococcusfaecium in the test product dropped from2 × 108 CFU/ml to about 104 to 105 CFU/ml.Since in vitro studies have shown that Strep-tococcus thermophilus is acid sensitiveand does not, to any significant degree, sur-vive passage through the small intestine(Agerbaek et al., 1995), it was suggestedthat it was the Enterococcus faecium thatplayed a significant role in the potentialbeneficial effect of this fermented milk prod-uct. Thus, a reduction in these bacteria mayhave been responsible for lowered efficacyof the product. In a subsequent study, Ses-sions et al. (1998) investigated the effects ofa similar milk product, fermented with thesame bacterial culture, on plasma choles-terol concentrations in men and women withmildly elevated cholesterol levels living inEngland. The study was multicentred, pla-cebo controlled, and double blind. In thisstudy, no effect on serum cholesterol levelswas observed either in the placebo or in the
test groups and the bacterial content of thefermented milk remained constant through-out the experiment. The authors speculatedthat variability in the results obtained withthe two studies was related to inherent dif-ferences between the populations studied,either in dietary habits or in the intestinalflora.
Two other studies with the Gaio® prod-uct have shown evidence for its beneficialeffects, but further investigation is neededto clarify some of the findings. Bertolami etal. (1999) studied the effect on the lipidprofile of this fermented milk product onpatients with mild to moderate primary hy-percholesterolemia. The study was prospec-tive, randomized, double-blinded, and pla-cebo controlled, with a crossover design.Thirty-two patients with ages ranging be-tween 36 and 65 years old were included inthe trial. Intake of the product lasted for 8weeks. It was concluded that the fermentedmilk (Gaio®) produced a small (average 5%)but statistically significant decrease (P =0.004) in total serum cholesterol. However,not all subjects responded to the product,and three of them even showed increasedcholesterol levels. A recent study (Larsen etal., 2000) was aimed at evaluating the ef-fects of the Gaio® product and two alterna-tive products on the serum cholesterol lev-els of overweight and obese subjects. Thestudy was randomized, parallel, double-blind, placebo- and compliance controlled,and performed over a period of 8 weeks.Seventy overweight and obese subjects (20males and 50 females), 18 to 55 years old,were randomly included and closely matchedinto five groups. Four of the groups weregiven 450 ml of fermented milk productsdaily with similar dietary composition. Oneof the groups consumed the test yoghurtGaio®; two groups consumed two new yo-ghurts with different bacterial cultures (onewas fermented with two strains ofS. thermophilus and two strains of L. acido-
Cri
tical
Rev
iew
s in
Bio
chem
istr
y an
d M
olec
ular
Bio
logy
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f C
alga
ry o
n 09
/10/
13Fo
r pe
rson
al u
se o
nly.
264
philus, and the other with two strains ofS. thermophilus and one strain of L. rhamno-sus); the fourth group consumed placeboyoghurt. The last group was given two pla-cebo tablets daily. When comparing all fivetreatment groups, and after adjusting forsmall changes in body weight, a significantdecrease in LDL-cholesterol was reportedin the Gaio® product group only (8.4%,P<0.05), after 8 weeks. This decrease wouldcorrespond to a reduction in the risk factorfor CHD of 20 to 30%, which is of clinicalrelevance. However, the decrease observedin LDL-cholesterol was significantly differ-ent only from groups consuming placeboyoghurt and pills. The reason for the ob-served hypocholesterolemic effect of theGaio® product was, once again, proposed tobe related to the CAUSIDO® bacterial cul-ture, particularly E. faecium. Although thespecific in vivo action of the strain remainsto be investigated, its ability to tolerate bileand assimilate cholesterol were suggestedto be responsible for the observed effect. Infact, Rossi et al. (1999) tested the ability ofE. faecium, L. acidophilus, L. jugurti,S. thermophilus, and L. delbrueckii to de-crease cholesterol in vitro and to grow inthe presence of bile salts. The study showedthat among the strains tested, E. faeciumand the mixture of E. faecium plus L. acido-philus produced the greater cholesterol re-duction in the medium after 24 h anaerobicincubation (53 and 65%, respectively).
In a study by Schaafsma et al. (1998),the effect of the Actimel Cholesterol Con-trol® yoghurt (Danone), fermented by Lac-tobacillus acidophilus and with fructo-oli-gosaccharides added, on blood lipids in adultmale volunteers with borderline elevatedlevels of serum total cholesterol was evalu-ated. The study was a randomized, placebo-controlled, double-blind, two-way crossovertrial with two treatment periods of 3 weeksseparated by a washout period of 1 week.During the treatment periods, 30 male sub-
jects consumed 125 ml of either test or ref-erence product three times daily as part oftheir habitual diet. The test product wasmilk fermented by yoghurt starters and Lac-tobacillus acidophilus, and containing 2.5%(w/v) fructo-oligosaccharides, 0.5% (w/v)vegetable oil and 0.5% (w/v) milk fat. Thereference product was a traditional yoghurt,containing 1% (w/v) milk fat. The authorsreported that consumption of the test prod-uct resulted in significantly lower values forserum total cholesterol (4.4%, P< 0.001)and LDL-cholesterol (5.4%, P< 0.005).However, they stated that beneficial effectsof the test product on serum cholesterolwere largely related to an increase of thisparameter during consumption of the refer-ence product. Also, it needs to be furtherevaluated whether the hypocholesterolemiceffect of the test product was attributable toeither the Lactobacillus acidophilus strain,the fructo-oligosaccharides, or both. Thepotential dose and time dependency of theobserved effect on serum cholesterol levelalso needs to be assessed.
De Roos et al. (1999), in a randomized,placebo-controlled parallel trial, attemptedto evaluate whether the intake of Lactoba-cillus acidophilus strain L-1 lowered serumcholesterol levels in healthy men and womenwith normal to borderline high-cholesterollevels. The 78 subjects used in this studyconsumed 500 ml of control yoghurt dailyfor 2 weeks. They were then randomly allo-cated to receive 500 ml per day of eithercontrol yoghurt or yoghurt enriched withLactobacillus acidophilus L-1 for another 6weeks. No significant reduction of serumcholesterol levels in the subjects consumingthe yoghurt enriched with L. acidophilus L-1was observed after the trial.
A recent human clinical trial was aimedat further studying the interrelationship be-tween the intake of probiotic yoghurt andthe concentration of cholesterol fractions(Schaarmann et al., 2001). In this study, 29
Cri
tical
Rev
iew
s in
Bio
chem
istr
y an
d M
olec
ular
Bio
logy
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f C
alga
ry o
n 09
/10/
13Fo
r pe
rson
al u
se o
nly.
265
healthy women consumed 300 g per day ofa probiotic yoghurt (Lactobacillus acido-philus and Bifidobacterium longum) after aperiod of eating standard yoghurt (Strepto-coccus thermophilus and Lactobacilluslactis). The volunteers were divided in anormocholesterolemic group (total choles-terol <250 mg/dL) and a hypercholester-olemic group (total cholesterol >250 mg/dL). The experiment consisted of three pe-riods (placebo, standard yoghurt, andprobiotic yoghurt), each lasting for 51 days.A decrease in the concentration of LDL-cholesterol and triacylglycerides after con-sumption of standard and probiotic yoghurtswas reported. A larger decrease was ob-served after intake of the probiotic yoghurtwhen compared with the standard yoghurtin the hypercholesterolemic group, but this,however, was not significant. Inversely, aincrease in the HDL cholesterol was ob-served with the probiotic yoghurt that re-sulted in a smaller atherogenic ratio (LDL/HDL-cholesterol). No significant differenceswere observed when comparing the normo-cholesterolemic and the hypercholester-olemic groups.
B. Mechanism(s) of Action
Existing evidence from human and ani-mal studies suggests a moderate cholesterol-lowering action of some fermented dairyproducts; however, the potential mechanismsfor this claimed effect remain to be clarified.
Several in vitro experiments with lacto-bacilli (Gilliland et al., 1985; Raˇsic et al.,1992; Noh et al., 1997) and bifidobacteria(Tahri et al., 1995, 1996) showed evidencefor the ability of some strains of these bac-teria to assimilate cholesterol in the pres-ence of bile acids.
Gilliland and co-workers in an in vitrostudy in 1985 found that certain Lactobacil-
lus acidophilus strains could remove cho-lesterol from a growth medium only in thepresence of bile and under anaerobic condi-tions. Because these conditions are expectedto occur in the intestine, the authors con-cluded that this should enable the organ-isms to assimilate at least part of the choles-terol ingested in the diet, thus making itunavailable for absorption into the blood.However, the ability to grow in the pres-ence of bile and to remove cholesterol fromlaboratory medium was found to vary con-siderably among L. acidophilus strains(Gilliland et al., 1985; Gilliland and Walker,1990; Walker and Gilliland, 1993). It wassuggested that an organism must be biletolerant to manifest cholesterol uptake inthe intestinal tract. However, no apparentcorrelation between bile tolerance and theability to assimilate cholesterol has beenobserved. The manifestation of the purportedassimilation action in vivo was tested usingyoung pigs (Gilliland et al., 1985). Pigs wereselected as the animal model because theirdigestive system, distribution of coronaryarteries, and atherosclerotic tendencies re-semble those of humans (Ratcliffe andLuginbuhl, 1971). The results of the studyshowed that feeding of pigs with L. acido-philus RP32, selected for its ability to growin the presence of bile and to remove cho-lesterol from a laboratory medium, signifi-cantly inhibited the increase in serum cho-lesterol when fed a high-lipid diet (P<0.05).The mechanism underlying this effect wasreferred to as cholesterol assimilation bythe L. acidophilus strain (Gilliland et al.,1985).
The purported assimilation of choles-terol by L. acidophilus was further supportedby the work of Raˇsic et al. (1992). Theseauthors reported that L. acidophilus pos-sessed a significantly greater cholesteroluptake ability than Streptococcus thermo-philus and a commercial yoghurt culture.However, variations in the ability of cul-
Cri
tical
Rev
iew
s in
Bio
chem
istr
y an
d M
olec
ular
Bio
logy
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f C
alga
ry o
n 09
/10/
13Fo
r pe
rson
al u
se o
nly.
266
tures of yoghurt bacteria and bifidobacteriato assimilate cholesterol were observed.
It has been suggested that the diverseresults obtained in different human studiesregarding the hypocholesterolemic effect ofyoghurt and other fermented milk productscan be, at least in part, be due to the differ-ent bacterial strains used in fermentation(Agerbaek et al., 1995). Viability of theingested bacterial strains in the human gut,and the ability to colonize the small intes-tine (where most of the cholesterol absorp-tion would take place), could ultimately beexpected to be of key importance for thiseffect.
Subsequent work, by Klaver and vander Meer (1993), studied the mechanism ofthe proposed assimilation of cholesterol byLactobacillus acidophilus and Bifidobac-terium bifidum. They concluded that removalof cholesterol from the culture medium byL. acidophilus RP32 and other species wasnot due to bacterial uptake of cholesterol,but rather could relate to co-precipitationwith deconjugated bile salts in an acidicenvironment. Deconjugated bile acids areless soluble and less likely to be absorbedfrom the intestinal lumen than conjugatedbile salts (Klaver and van der Meer, 1993;De Rodas et al., 1996). Thus, deconjugationof bile acids in the small intestine couldresult in a greater excretion of bile acidsfrom the intestinal tract, especially as freebile acids are excreted more rapidly thantheir conjugated forms. Increased excretionof bile acids should result in lowered serumconcentrations, which in turn would decreasethe amount of bile acids reaching the liverfor secretion back into the intestine throughenterohepatic circulation. To replace theexcreted bile acids, more would have to besynthesized from cholesterol in the liver.Thus, it has been suggested that in a steady-state situation, deconjugation of bile acidscould lead to the reduction of serum choles-terol by increasing the formation of new
bile acids or by reducing the absorption ofcholesterol throughout the intestinal lumen.However, it remains unclear whether mi-crobial bile deconjugation properties arerelated to the hypocholesterolemic effectsobserved in vivo (Gilliland et al., 1985),because pH in the intestinal tract of humansis usually neutral to alkaline.
Tahri et al. (1995) studied the hypoth-esis of the proposed assimilation orcoprecipitation of cholesterol by Bifidobac-terium species. They observed the existenceof an intense binding between cell surfaceand cholesterol, which was considered as anuptake of cholesterol into the cells. Thisassimilation was dependent on cell growthand the presence of bile salts was shown tobe a prerequisite for significant cholesterolremoval. The authors concluded that theobserved removal of cholesterol from thebroth could not be attributed only to thecoprecipitation of cholesterol with decon-jugated bile salts, but rather to a conjuga-tion of both effects. It still remains neces-sary to determine whether the requirementfor bile salts is related to their detergenteffect and if an inhibition of bile saltdeconjugation influences the assimilationof cholesterol. Follow-up work (Tahri et al.,1996) further supported this mechanism. Theresults obtained in this study confirmed thework of Klaver and van der Meer (1993), inthat, in the absence of bacterial cells, cho-lesterol was partially removed from themedium at pH values lower than 5.5 whendeconjugated bile salts were added. How-ever, the authors showed that this was atransient phenomenon because the precipi-tated cholesterol was redissolved when so-lutions were adjusted to pH 7. Furthermore,they concluded that resting cells ofbifidobacteria do not interact with choles-terol, while growing cells of bifidobacteriado seem to assimilate cholesterol in theircell membrane. No significant relationshipwas, however, noticed between cholesterol
Cri
tical
Rev
iew
s in
Bio
chem
istr
y an
d M
olec
ular
Bio
logy
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f C
alga
ry o
n 09
/10/
13Fo
r pe
rson
al u
se o
nly.
267
assimilation and the degree of bile saltdeconjugation by bifidobacteria, as previ-ously reported for lactobacilli (Gilliland etal., 1985). It has been suggested that bilesalts hydrolase activity might have a role inthe mechanism of assimilation as nondecon-jugating bifidobacteria were not able to as-similate appreciable amounts of cholesteroleven if conjugated or deconjugated formsof bile salts were present in the growthmedium. The authors concluded that thischolesterol assimilation model could explainthe in vivo observed hypocholesterolemiceffect. Further research is still needed todetermine the mechanism of cholesterolassimilation and localisation of the assimi-lated cholesterol in the cells.
In an animal study designed to evaluatethe hypocholesterolemic effect of L. acido-philus ATCC43121, 33 pigs were fed a highcholesterol diet for 14 days (De Rodas et al.,1996). Following this, on day 15 and for anadditional 15 days, the pigs were assignedto one of four treatment groups, includingtwo levels of calcium (0.7 and 1.4%), withor without the L. acidophilus strain added(2.5 × 1011 CFU/ml). It was reported thatpigs fed L. acidophilus had 11.8% lowertotal cholesterol than pigs fed a diet withoutL. acidophilus. Similarly, pigs fed 1.4% cal-cium had significantly lower total choles-terol than pigs fed 0.7% calcium. In addi-tion, during the overall 15-day period, areduction of 23.9% in serum bile acids bydietary L. acidophilus and by 21.4% by 1.4%dietary calcium compared with the controlswas observed. Based on these results, theauthors concluded that both L. acidophilusATCC43121 and calcium could enhance thereduction of serum cholesterol in pigs thathad been fed a high-cholesterol diet. Theysuggested that with regards to dietaryL. acidophilus, this is probably done throughalterations in the enterohepatic circulationof bile acids as the greater decline in serumcholesterol, coupled with lower serum bile
salts could indicate an association betweenthe deconjugation action of the strain and itshypocholesterolemic effect. Similarly, theexact mechanism for the LDL-cholesterollowering effect of calcium is still unclear. Ithas been hypothesized that excess dietarycalcium can bind bile acids and thereforesuppresses reabsorption into the enterohe-patic circulation.
Further evidence for cholesterol assimi-lation by L. acidophilus ATCC43121 wasgiven by Noh et al. (1997). These authorsreported that cholesterol assimilated by thisstrain was not metabolically degraded, asmost of it was recovered with the cells.They observed that cells grown in the pres-ence of cholesterol micelles and bile saltswere more resistant to lysis by sonicationthan were those grown in their absence.This could suggest a possible alteration ofthe cell wall or membrane by cholesterolsuch that lactobacilli were more resistant tosonic disruption. The in vivo effect of aci-dophilus yoghurt on serum cholesterol lev-els was evaluated using mice as a live modelby Akalin et al, (1997). In this study, 60male mice were assigned to three dietarytreatments for 56 days: water (control), yo-ghurt made from milk inoculated withS. thermophilus and L. delbrueckii, and yo-ghurt made from milk inoculated withS. thermophilus and L. acidophilus. A sig-nificant decrease in the mean values fortotal and LDL-cholesterol concentrationswas observed in the group given acidophi-lus yoghurt on days 28 and 56. The hypo-cholesterolemic effect of ‘standard’ yoghurtoccurred only at the end of the 56 d periodand was lower than that of acidophilus yo-ghurt. Both HDL-cholesterol and triglycer-ide concentrations seemed to be unaffectedby dietary yoghurt or acidophilus yoghurt.
A recent study (Meei-YN Lin, 2000)investigated the in vitro cholesterol reduc-ing abilities of six L. acidophilus strains(including Nestle LC1® strain La1). A maxi-
Cri
tical
Rev
iew
s in
Bio
chem
istr
y an
d M
olec
ular
Bio
logy
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f C
alga
ry o
n 09
/10/
13Fo
r pe
rson
al u
se o
nly.
268
mum cholesterol uptake of 57% was re-ported when L. acidophilus ATCC4356 wasgrown anaerobically for 24 h in a mediumsupplemented with bile acids (oxgall). Theauthors concluded that the in vivo hypo-cholesterolemic effect of L. acidophilus cellswas due to the direct assimilation by thecells and/or attachment of cholesterol to theirsurface. As previously mentioned, copre-cipitation of cholesterol with decon-jugatedbile acids, which occurs in vitro, is not likelyto take place in vivo because pH of the smallintestine is higher than 6.0.
Lactic acid bacteria with active BSH, orcultured products containing them, are sug-gested to lower serum cholesterol levelsthrough an interaction with the host bile saltmetabolism (De Smet et al., 1994). The pro-posed mechanism of cholesterol reductionis comparable to that of a cholestyraminetreatment, which, like other bile saltsequestrants such as colestipol, binds bilesalts and prevents them from being reab-sorbed. Thus, less bile salts would return tothe liver, resulting in a loss of feedbackinhibition of bile salt synthesis and an in-creased conversion of cholesterol to bilesalts. This led the authors to suggest that theingestion of lactic acid bacteria containingactive BSH might be regarded as a ‘biologi-cal’ alternative to common medical or sur-gical interventions to treat hypercholester-olemia. Based on these assumptions, DeSmet et al. (1998) studied the effects offeeding live Lactobacillus reuteri cells con-taining active bile salt hydrolase (BSH) onplasma cholesterol levels in pigs. The studylasted for 13 weeks. During the treatmentperiod (4 weeks), 20 pigs were given, twicedaily, 1011,25 CFU of the strain. It was re-ported that after 2 weeks of Lactobacillusreuteri feeding, total and LDL-cholesterolin treated pigs were reduced by 11 and 26%,respectively, compared with control animals(P < 0.05). After 4 weeks these reductionswere 15 and 24%, respectively (P < 0.05).
No change was observed in HDL-choles-terol concentrations. During the subsequentposttreatment follow-up period, it was re-ported that total and LDL-cholesterol con-centrations in treated pigs increased, butthese levels were still significantly lowerthan values measured in control pigs. Arecent animal study involving L. reuteri wasaimed at an evaluation of the effect of ad-ministration of L. reuteri CRL1098 (104 cfu/day) to mice for 7 days prior to inducinghypercholesterolemia (Taranto et al., 2000).Even at this low dose, L. reuteri seemed tobe effective in preventing hypercholester-olemia in mice, producing a 17% increasein the ratio of HDL/LDL lipoprotein. De-creases in total cholesterol and triglyceridelevels of 22 and 33%, respectively, wereobserved in the group given milk withL. reuteri in comparison with the group givenmilk alone. Additionally, serum cholesterolin the L. reuteri pretreated mice increasedby only 38% compared with the 82% in-crease observed in the group given milk. Itwas concluded that L. reuteri might be ef-fective as a prophylactic agent in prevent-ing hypercholesterolemia.
In a recent in vitro study, Usman andHosono (1999) noticed a significant varia-tion among 28 strains of Lactobacillusgasseri in their growth performance in me-dium containing bile, as well as in theirability to bind cholesterol. They further sug-gested that no significant correlation ex-isted between bile tolerance and the choles-terol binding ability. In this study, all strainsof human origin tested showed greater varia-tion in bile tolerance in agreement with pre-vious work (Gilliland et al., 1985; Klaverand van der Meer, 1993). It was suggestedthat differences in bile tolerance could bedue to growth performance of the individualstrains. While in previous studies the actionof L. acidophilus on cholesterol uptake wasevaluated when cells were grown in MRSbroth containing a certain amount of bile
Cri
tical
Rev
iew
s in
Bio
chem
istr
y an
d M
olec
ular
Bio
logy
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f C
alga
ry o
n 09
/10/
13Fo
r pe
rson
al u
se o
nly.
269
salts and cholesterol, the study of Usmanand Hosono (1999) was focused on the bind-ing abilities of the intact cells toward cho-lesterol. The authors found that all 28 strainswere able to bind cholesterol, but the bind-ing abilities varied widely. It was suggestedthat differences were related to chemicaland structural properties of the bacterial cell-wall peptidoglycans. However, it remainsunclear whether the variation in cholesteroluptake by the strains of L. gasseri was dueto difference in their cell membranes or toother cell components. In a follow-up tothis study, Usman and Hosono (2000) in-vestigated the effect of a nonfermented milksupplemented with L. gasseri (109 cfu/ml)on serum lipids and fecal steroids in hyper-cholesterolemic rats. Both strains used inthis study (SBT0270 and SBT0274) havepreviously been shown to exhibit high cho-lesterol-binding and taurocholate-deconju-gating activities in vitro (Usman and Hosono,1999). The total and LDL-cholesterol levelswere reported to be 42 and 64%, respec-tively, lower in the group given L. gasseriSBT0270 supplemented milk than in themilk-only group. Triglyceride levels de-creased when rats were given milk andnonfermented milk supplemented with thebacterial strains in relation to the controlgroup (water), while milk alone had no ob-served effect on total and LDL-cholesterol.From the results obtained in this study, theauthors attributed the observed hypo-cholesterolemic effect of L. gasseriSBT0270 to its ability to suppress the reab-sorption of bile acids into the enterohepaticcirculation (by deconjugation) and to en-hance the excretion of acidic steroids infeces.
A recent animal study involving L. caseistrain Shirota (Yakult®) was aimed at deter-mining the effect of skim milk fermented bythis strain on plasma lipids in hamsters(Kikuchi-Hayakawa et al., 2000). A 30%decrease in the triglyceride plasma levels
was observed in hamsters fed the enricheddiet containing the fermented milk product(FMP) compared with the control diet (un-fermented milk). However, plasma choles-terol concentration was not affected by theFMP supplement to the diet. In fact in an invitro experiment the same authors observedthat L. casei strain Shirota, despite growingwell in the presence of mixed lipid micellescontaining bile acids, did not have the abil-ity to significantly remove cholesterol fromthe culture broth after 24 h anaerobic incu-bation (11%), contrary to some other strainstested (L. acidophilus [max. 83%],L. crispatus [max. 83%], L. gasseri [max.80%]).
Probiotic bacteria once resident in thehuman gut ferment food-derived indigest-ible carbohydrates that results in an increasedproduction of short-chain fatty acids(SCFA). This has been suggested to cause adecrease in the systemic levels of bloodlipids either by inhibiting hepatic choles-terol synthesis (to an unknown extent), orby redistributing cholesterol from plasma tothe liver (Fuller and Gibson, 1998). TheSCFA production in the large intestine hasbeen reported to be 100 to 450 mmol/day,with relative proportions of acetate, propi-onate, and butyrate being about 60:20:15,depending on the substrate (St-Onge et al.,2000). Acetate in the serum seems to in-crease total cholesterol, while propionateincreases blood glucose and tends to lowerthe hypocholesterolemic response caused byacetate, by reducing its utilization by theliver for fatty acid and cholesterol synthe-sis. Indeed, a study by Wolever and col-leagues (1996) showed that serum acetateand propionate concentrations were relatedto serum lipid concentrations in both malesand females, serum propionate beingstrongly negatively related to both total andLDL cholesterol (P < 0.001). However, suf-ficient propionate must be produced to off-set the effects of acetate generation as a
Cri
tical
Rev
iew
s in
Bio
chem
istr
y an
d M
olec
ular
Bio
logy
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f C
alga
ry o
n 09
/10/
13Fo
r pe
rson
al u
se o
nly.
270
precursor for lipid synthesis. Therefore,depending on the proportion of each fattyacid produced during bacterial fermenta-tion, plasma cholesterol concentrations thusmay be altered through this mechanism.
Additionally, some bacteria may inter-fere with cholesterol absorption from thegut either by deconjugating bile salts andtherefore affecting the metabolism of cho-lesterol or by directly assimilating choles-terol. A number of bacteria have been re-ported to hydrolyze conjugated bile acids,such as Bacteroides spp., bifidobacteria,fusobacteria, clostridia, lactobacilli, andstreptococci (Hylemon and Glass, 1983).There is evidence that the gut flora not onlyhydrogenates, dehydrogenates, and oxidizesbile acids, but also cleave side chains toyield steroids. In vivo, removal of choles-terol would occur because deconjugated bileacids are not well absorbed by the gut mu-cosa and are excreted through the faecesand urine. The excretion of bile acids re-sults in decreased enterohepatic recircula-tion and therefore more cholesterol, whichis the precursor of bile acids, needs to beutilized for de novo bile acid synthesis.
As mentioned previously, in vitro stud-ies using different strains of L. acidophilusgrown on media containing bile have shownthat certain strains can modify cholesterolmetabolism. Because the amount of bile inthe medium did not exceed concentrationsnormally found in the intestine, it can beexpected that this cholesterol assimilationwould also occur in vivo. The uptake ofcholesterol by bacteria would make it un-available for absorption into the circulation.
However, some work still needs to beundertaken in this field leading toward moreconclusive clinical evaluations to under-standing of the in vivo mechanisms ofprobiotic effect on blood lipids and to im-provement of strain stability characteristics.Also, the in vitro ability to reduce choles-terol observed with some strains needs to be
confirmed in mixed culture and mixed sub-strate environments.
III. PREBIOTICS
As previously mentioned, evidence ex-ists that alterations in gut microflora in-duced by fermentation of nondigestible com-ponents of diet (prebiotics) may have thepotential to influence systemic lipid me-tabolism. This has attracted considerableattention, mainly because unlike probiotics,prebiotic substances by definition are notsubject to viability problems. Such sub-stances also have greater possibilities forincorporation into a wide range of commonfoodstuffs.
In contrast with the use of live microor-ganisms in foods, a prebiotic supplementtargets certain components of the indigenousgut microbiota (Gibson and Roberfroid,1995). The usual target organisms arebifidobacteria and lactobacilli, as they arenatural residents of the gut microbiota. Theprebiotic approach advocates use of nonvi-able food components that remain unab-sorbed and nondegraded in the upper gas-trointestinal tract, but have a selectivefermentation in the large gut, exploiting thefact that diet is a major determinant of co-lonic microflora function (Gibson, 1999).Any food that reaches the colon such asnondigestible carbohydrates, some peptidesand proteins, as well as certain lipids, is apotential prebiotic (Gibson, 1999). Nondi-gestible carbohydrates, in particular fruc-tose oligosaccharides, are the most studiedprebiotics. Fructooligosaccharides (FOS)consist of short- and medium-length chainsof β-D-fructans in which fructosyl units arebound by a β 2-1 linkage, with the degree ofpolymerization varying between 2 and 60(inulin) or 2 and 20 (oligofructose) (Gibson,1999). FOS such as inulin and oligofructose
Cri
tical
Rev
iew
s in
Bio
chem
istr
y an
d M
olec
ular
Bio
logy
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f C
alga
ry o
n 09
/10/
13Fo
r pe
rson
al u
se o
nly.
271
are naturally occurring indigestible carbo-hydrates found in many plant foods likechicory, onions, and asparagus (Roberfroid,1996). These compounds escape hydrolysisand absorption in the human upper diges-tive tract and reach the colon where they arefermented by the intestinal bacteria, espe-cially bifidobacteria (Gibson and Roberfroid,1995; Roberfroid, 1996), to SCFA (mainlyacetate) and other metabolites (e.g., lactate;Gibson, 1999).
A. Human and Animal DietaryStudies
Lipid-lowering effects of dietary oli-gosaccharides have been proposed and spe-cifically a decrease in the plasma triacyl-glycerol concentration was observed (Levratet al., 1991; Delzenne et al., 1993; Fiordalisoet al., 1995).
Fiordaliso et al. (1995) showed that theadministration of oligofructose decreasedtotal cholesterol levels in the serum of rats.Oligofructose supplementation has also beenshown to protect rats against the increase infree cholesterol concentration induced byhigh-fat diets, without preventing an accu-mulation of cholesterol in liver tissue (Koket al., 1996).
Delzenne et al. (1993) studied the influ-ence of dietary FOSs on lipid metabolism inrats. In this experiment the animals werefed oligofructose in the diet for 30 days, ata dose of 20 g/100 g. A large decrease in theconcentration of triglycerides, both in theserum and liver, was reported in test ani-mals compared with the levels measured inisocalorically fed control rats. The amountof total cholesterol did not change, but anincrease in the ratio of HDL/LDL lipopro-teins was observed. The same pattern ofseric and hepatic lipid modifications hasbeen observed in rats fed 10% inulin by
weight (Levrat et al., 1991). The existenceof a putative role of some fermentation prod-ucts (e.g., SCFA) as mediators of the sys-temic effects of FOSs was hypothesized(Delzenne et al., 1993). Also, Levrat et al.(1991) showed that high propionic acid fer-mentations were present in the cecum ofrats fed diets containing moderate propor-tions of inulin. The authors proposed thatthe hypocholesterolemic effect of inulincorresponded to a metabolic effect, becausethe oligosaccharide structure of inulin wouldbe unsuitable for adsorption of neutral oracidic steroids in the small intestine.
In a subsequent study, normochol-esterolemic rats were fed a bread diet con-taining either corn starch or 6%(w/v) inulin, and the diets were either cho-lesterol-free or contained 1% (w/v) choles-terol and 0.1% (w/v) cholic acid (Vanhoofand Schrijver, 1995). A significant reduc-tion in plasma cholesterol concentrations inthe cholesterol-free diets with added inulinand a tendency to elevated fecal excretionof neutral steroids were reported. The au-thors suggested that the increase in choles-terol excretion in animals receiving inulinmight be caused by a decrease in choles-terol absorption through a higher viscosityin the upper intestinal tract. This would re-sult in a higher cholesterol catabolism in theliver, which would lead to lower plasmacholesterol concentrations. Indeed, an in-verse linear relationship between liver cho-lesterol concentrations and daily fecal ex-cretion of bile acids was observed in testanimals. It was concluded that the choles-terol-lowering effect of inulin in normo-cholesterolemic rats might be due to in-creased fecal neutral steroid and bile acidexcretion. The suggested mechanism for thiseffect was a reduction in cecal pH observedin rats consuming inulin (Vanhoof andSchrijver, 1995). At a lower pH the amountof soluble bile acids decreases, as a resultlipid absorption decreases and fecal bile acid
Cri
tical
Rev
iew
s in
Bio
chem
istr
y an
d M
olec
ular
Bio
logy
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f C
alga
ry o
n 09
/10/
13Fo
r pe
rson
al u
se o
nly.
272
excretion increases, despite the fact thatinulin itself is unlikely to bind bile acids inthe upper digestive tract (Levrat et al., 1994).As previously mentioned, higher fecal bileacid excretion leads to increased utilizationof liver cholesterol to resynthesize bile ac-ids (Klaver and van der Meer, 1993;De Smet et al., 1994). In a similar experi-ment with hypercholesterolemic rats, a ten-dency to higher bile acid excretion was re-ported, but no increase in total neutral steroidexcretion and no significant effect on plasmaand liver cholesterol concentrations wasnoted (Vanhoof and Schrijver, 1995).
The accurate differences in the resultsobtained with normo- and hypercholester-olemic rats make conclusions difficult. Fur-ther studies are required to elucidate themechanisms by which inulin may elicit itslipid-lowering effect in humans, because lipidmetabolism is somewhat different from therat models used. In an attempt to overcomethis, Trautwein et al. (1998) used Syrian ham-sters, which presumably provide a bettermodel for the study of lipid and cholesterolmetabolism than rats. This experiment wasdesigned to further evaluate the lipid-lower-ing potential of inulin and especially its ef-fect on bile acid metabolism. The animalswere fed diets containing different amountsof inulin (8, 12, and 16%) for 5 weeks. Theauthors reported that addition of inulin to thecholesterol-enriched diet of test animals ex-erted significant hypocholesterolemic andhypotriacylglycerolemic effects, especially atdietary levels of 12 and 16%. They observedthat all three levels of dietary inulin causeddistinct alterations in the bile acid profile ofgallbladder bile. It was concluded that thelipid-lowering action of inulin was possiblydue to several mechanisms, including alteredhepatic triacylglycerol synthesis and VLDLsecretion, and impaired reabsorption of cir-culating bile acids. In a recent study byDaubioul and colleagues (2000), the effectsof oligofructose on lipid metabolism in obese
Zucker rats was examined. It was observedthat the addition of oligofructose to the dietslowed the increase in body weight withoutmodifying serum triglycerides or glucoseconcentrations after 7 weeks of treatment.However, after 10 weeks a 57% decrease inthe hepatic concentration of triglycerides,relative to controls, was reported. The influ-ence of dietary oligofructose on postprandialtriglyceridemia, an important variable asso-ciated with development of atherosclerosisin humans, remains to be further clarified.
Because animal studies have shownsome evidence for the lipid-lowering ef-fects of FOSs, much attention has been givento their potential effect in humans. The num-ber of human studies is, however, limitedand results concerning effects on plasmacholesterol and triacylglycerol are not con-clusive (Table 2).
Yamashita et al. (1984) studied the ef-fect of oligofructose intake on blood lipidlevels of individuals with noninsulin-depen-dent diabetes. They reported a 8% reductionin total, and 10% reduction in LDL-choles-terol after the administration of 8 g of asynthetic oligosaccharide for 14 days, com-pared with a control group given sucrose inthe same food vehicles. They also observedthat the reduction was greater in hypercho-lesterolemic subjects. No effect on circulat-ing triacylglycerols was reported. This studywas conducted in a parallel design with anintervention and a control group, and conse-quently had lower statistical power due tothe lack of a crossover design where sub-jects serve as their own control.
More recently, Canzi et al. (1995) stud-ied the effect on lipid metabolism of a pro-longed ingestion of 9 g/day of inulin fromchicory incorporated into a ready-to-eatbreakfast cereal in 12 healthy young malevolunteers. A marked effect of inulin inreducing fasting triglycerides concentrations(–27%), and to a lesser extent total choles-terol (–5%), without undue effects on HDL-
Cri
tical
Rev
iew
s in
Bio
chem
istr
y an
d M
olec
ular
Bio
logy
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f C
alga
ry o
n 09
/10/
13Fo
r pe
rson
al u
se o
nly.
273
TABLE 2Summary of Human Dietary Studies on the Effect of Fructooligosaccharides(FOS) on Lipid Metabolism
Cri
tical
Rev
iew
s in
Bio
chem
istr
y an
d M
olec
ular
Bio
logy
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f C
alga
ry o
n 09
/10/
13Fo
r pe
rson
al u
se o
nly.
274
cholesterol, was reported. From this study,it was concluded that the effects of inulin onplasma cholesterol were not related to intes-tinal microbial metabolism of bile acids, ashas been previously suggested (Yamashitaet al., 1984), because they observed thatinulin consumption did not affect the countsof bile acids dehydroxylating microorgan-isms. The authors also observed a large andconsistent rise in breath hydrogen, indicat-ing the possibility that some other microbialactivities may have been changed in theproximal colon despite the presence of astable microflora.
In a subsequent study, the effect of inulinand oligofructose on cholesterol absorptionand excretion in humans was evaluated fur-ther (Ellegard et al., 1997). This study wasbased on a double blind crossover procedureand included 10 ileostomy subjects. The sub-jects were given diets, including inulin,oligofructose, or sucrose during three experi-mental periods of 3 days each. Between thedietary periods there was a washout period of4 days. No difference was noticed by thesubjects between the three diets, which indi-cates that the doses given (17 g of inulin andoligofructose) were well tolerated. No sig-nificant effect of inulin or oligofructose onabsorption of cholesterol, or excretion of fat,cholesterol, or bile acids was observed dur-ing the study. This led the authors to con-clude that the systemic effects of inulin andoligofructose reported previously were notmediated via increases in cholesterol and ste-rol excretion (the proposed explanation forthe hypocholesterolemic effect of othersoluble dietary fibers such as pectins;Andersson, 1993).
In another human dietary study, Pedersenet al. (1997) evaluated the effect of a dailyintake of 14 g inulin, added to a low-fatspread, on fasting blood lipids in 64 youngwomen in a randomized, double-blind, cross-over trial involving two periods of 4 weeks.No significant differences in plasma total
cholesterol, HDL- and LDL-cholesterol andtriacylglycerol concentrations were observedbetween the placebo and inulin periods. How-ever, the authors reported a significant de-crease in the LDL-cholesterol:HDL-choles-terol ratio at the end of both the control periodand the inulin period. These results were inaccordance with those reported in a previousstudy (Luo et al., 1996) involving 12 younghealthy males who ingested 20 g FOS/dayfor 4 weeks in a randomized, crossover ex-periment.
A follow-up study examined the effectof dietary inulin on serum lipids in hyperc-holesterolemic subjects (Davidson et al.,1998). This was a randomized, double-blind,crossover study performed using 21 adultswith mild-to-moderate hypercholester-olemia, with two 6-week treatment periodsseparated by a 6-week washout. During thetreatment periods the subjects consumedthree servings per day of inulin-containingfoods, corresponding to a total of 18 g/day.Significant reductions were reported whencomparing the response between periods,either in LDL-cholesterol (–14.4%, P<0.05)or in total cholesterol (–8.7%, P<0.05). How-ever, when comparing the values for any ofthe lipid variables at the end of the inulinand control periods, no differences wereobserved. This was mainly attributed to asignificant increase in total cholesterol andLDL-cholesterol observed during the con-trol phase, whereas these values did notchange appreciably during the inulin phase.Although it was not possible to draw firmconclusions, based on previously reporteddata showing a lipid-lowering effect of inu-lin, the authors suggested that inulin con-sumption prevented the increase in total andLDL-cholesterol observed during the con-trol period.
In a recently conducted study, 54 middle-aged subjects with moderately raised bloodlipid concentrations, consumed 10 g per dayof inulin or placebo, in a powdered form
Cri
tical
Rev
iew
s in
Bio
chem
istr
y an
d M
olec
ular
Bio
logy
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f C
alga
ry o
n 09
/10/
13Fo
r pe
rson
al u
se o
nly.
275
(Williams, 1999). No significant changes inconcentrations of total, LDL-, or HDL-cho-lesterol were seen in either group over the8-week intervention. In contrast, it was re-ported that fasting serum triglycerides weresignificantly lower (19%) after 8 weeks inthe inulin-treated group, returning to baselinevalues 4 weeks after treatment. Similar re-sults were reported by Causey et al. (2000)in a dietary study involving hypercholester-olemic men. In this study 12 male subjectswere randomly assigned to two controlleddiets that differed only in that the test dietcontained a daily intake of 20 g of inulin,while in the control diet that was replacedby sucrose, and consumed each diet for 3weeks. A significant decrease in serumtriglycerides levels (40 mg/dL, 14%) wasobserved with the inulin diet. A trend to-ward a reduction in serum cholesterol wasalso observed, although this was not signifi-cant.
A recent meta-analysis on the cholesterol-lowering effects of dietary fiber suggested thatsoluble fibers (pectin, oat bran, guar gum, andpsyllium) had a small but significant decreas-ing effect on total and LDL-cholesterol levelswithin the practical range of intake (Brown etal., 1999). Inulin appears to have a similareffect on blood lipids when consumed byhyperlipidemic adults. Therefore, preliminaryevidence exists for a hypotriglyceridemic ef-fect of FOSs, but, at the present stage of knowl-edge, it is not possible to conclude ahypocholesterolemic effect.
B. Mechanism(s) of Action
Oligosaccharides are rapidly and com-pletely fermented in the colon (Gibson, 1999)and one hypothesis for a lipid-lowering mecha-nism is by increasing the synthesis of fermen-tation byproducts (e.g., propionate) which reachthe liver by the portal vein and potentially
modulate the hepatic cholesterol synthesis (Chenet al.,1984; Levrat et al., 1994). This hypothesiswas supported by a study in rats showing thatthe decrease in blood lipids after daily admin-istration of an oligofructose-rich diet was asso-ciated with a reduction in plasma VLDL par-ticles, and a reduced capacity of isolatedhepatocytes to synthesize triacylglycerols(Fiordaliso et al., 1995). Although it was dem-onstrated that oligosaccharides induce high pro-pionic fermentations in the cecum of rats (Levratet al., 1994), this fact is considered controver-sial and does not seem to play a major role inthe hypocholesterolemic effect of FOSs. More-over, several animal studies have reported areduction in intestinal cholesterol and bile acidabsorption, with a consequent increase in bileacid and neutral steroid excretion (Vanhoof etal., 1995; Trautwein et al., 1998). Such findingssuggest that interruption of the enterohepaticcirculation of bile acids with an enhanced fecalexcretion could have a major impact on theobserved hypocholesterolemic effect.
Evidence exists that other mecha-nisms might also be involved in thehypocholesterolemic action of FOSs, butthere are still not enough data to drawdefinitive conclusions.
In contrast, it is commonly accepted thatthe principal mechanism for the hypo-triacylglycerolemic effect of oligofructose andinulin is a reduction in the hepatic de novo fattyacid and triacylglycerol synthesis (Fiordaliso etal., 1995; Kok et al., 1996). Data from thesestudies presented evidence for the hypothesisof a decreased de novo lipogenesis in the liver,through a coordinated reduction in the activityof all lipogenic enzymes. This hypothesis wasalso supported by a reduction in plasma VLDLparticles observed in the study by Trautwein etal. (1998), indicating a decreased productionand secretion of VLDL-triglyceride. Furtherevidence for this mechanism was given by re-cently published work aimed at reviewing theprobable biochemical mechanism(s) account-ing for the hypolipidemic effect of FOSs
Cri
tical
Rev
iew
s in
Bio
chem
istr
y an
d M
olec
ular
Bio
logy
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f C
alga
ry o
n 09
/10/
13Fo
r pe
rson
al u
se o
nly.
276
(Delzenne and Kok, 1999). These authors re-ported a significant postprandial triglyceridelowering effect (27 to 61%, P < 0.05) afteradministrating 10 g/100 g oligofructose to maleWistar rats fed either a standard diet, a fiber-free diet or a high-fat diet. The authors con-cluded that the triacylglycerol-lowering effectof oligofructose observed in rats was caused byits antilipogenic action in the liver, that is, byreducing the activity and possibly the expres-sion of all lipogenic enzymes.
It has also been observed that FOS in-take significantly reduces serum insulin andglucose (Kok et al., 1996), which are bothknown to be present in the nutritional regu-lation of lipogenesis. FOS also increases theproduction of intestinal peptides, namely,GIP and GLP-1. Both of these peptides areknown to regulate postprandial insulin re-lease and also to have direct insulin-likeactions on lipid metabolism (Morgan, 1996).In conclusion, a relationship seems to existbetween the modulation of hormone andintestinal peptide production and theantilipogenic effect of FOSs. However, theexact mechanism of action needs to be clari-fied.
IV. CONCLUSION
In conclusion, dairy products fermentedwith the appropriate strain(s) of bacteria canbe anticipated to induce a lowering of circu-lating cholesterol concentrations, thus dimin-ishing the risk of CHD. However, the strainsof bacteria used must be bile tolerant, havethe ability to deconjugate bile acids, and bindcholesterol. If these criteria are fulfilled, fer-mented dairy products can be viewed as ‘func-tional’ foods in lowering elevated choles-terol concentrations. There are a number ofmechanisms suggested for the purported cho-lesterol-lowering action of probiotics. Theseinclude physiological actions of the end prod-
ucts of SCFA fermentation, cholesterol as-similation by the probiotic bacteria, choles-terol binding to the bacterial cell wall, andenzymatic deconjugation of bile acids. Thedeconjugation of bile acids by gut bacteriaincreases their rates of excretion and there-fore would lead to a higher cholesterol de-mand by the liver for de novo synthesis ofbile acids, with a consequent reduction onserum cholesterol. One concern about thismechanism was suggested to be the potentialincreased risk for colon cancer due to thecarcinogenic properties of the deconjugatedbile acids (Sanders, 2000). These hypothesesneed to be confirmed in animal and humanstudies and the exact mechanism(s) of actionof probiotic bacteria on cholesterol reductionremain unclear.
To date, products that contain live bac-teria, such as yoghurt, acidophilus milk, andkefir (fermented dairy products containingseveral types of bacteria in symbiosis withyeasts) contain strains that do not normallyreside in the human intestinal tract. Theresult being that these bacteria are not ableto colonize the intestine and are quicklyeliminated in faeces. As such, daily con-sumption of probiotic products is necessaryfor any long-term effect on metabolism. Thisfact could account for some of the conflict-ing results obtained from several humanand animal dietary studies.
In what concerns prebiotics, the dataavailable at present are still rather inconsis-tent, but seem to indicate that intake ofmoderate levels of inulin or oligofructosemay affect, to some extent, human lipidmetabolism. Although convincing lipid-low-ering effects of inulin and oligofructose havebeen observed in animals, the studies haveused relatively high dose levels (50 to 200g/kg). Equivalent doses could not be used inhumans because of known adverse gas-trointestinal side effects at intake levelsabove 30 g/day (Delzenne and Kok, 1999).Studies that have investigated the effects of
Cri
tical
Rev
iew
s in
Bio
chem
istr
y an
d M
olec
ular
Bio
logy
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f C
alga
ry o
n 09
/10/
13Fo
r pe
rson
al u
se o
nly.
277
inulin and oligofructose in humans are fewin number, although those that have beenconducted are well designed and includerelatively large numbers of subjects. Fur-ther research is required to confirm the find-ings. In studies conducted in normal-lipidemic subjects, two reported no effectsof inulin or oligofructose on serum lipids,whereas two others reported significant re-ductions in serum triglycerides (–19 and–27%), with more modest changes in serumtotal and LDL cholesterol. Present data sug-gest that in subjects with hyperlipidemia,any effects that do occur result primarily inreductions in cholesterol, whereas in nor-mal subjects effects on serum triglyceridesare the dominant feature. This latter responseis similar to that observed in animals inwhich the observed effects on cholesterolare small. This led some authors to con-clude that the hypocholesterolemic actionof FOSs is modified by dietary cholesterol(Fernandez, 1995), and therefore thehypocholesterolemic response in humansubjects could possibly vary depending onwhether individuals are normal or hyperc-holesterolemic.
As mentioned previously, animal stud-ies have identified the inhibition of hepaticlipogenesis as the major site of action forthe triglyceride-lowering effects of FOS, andthis pathway is known to be relatively inac-tive in humans unless a high-carbohydratediet is fed (Delzenne and Kok, 1999). Assuch, one could suggest that the apparentlack of effect observed in some human stud-ies performed in healthy humans (whosediet consists of much less carbohydrates,and more lipids than rodents) is not surpris-ing and does not demonstrate an absence ofeffect. Future attempts to demonstrate lipid-lowering effects of FOSs intake must con-sider the nature of the background diet as adeterminant of response.
Therefore, despite the great interestit could constitute in terms of human
health, the mechanisms of the effects ofnondigestible, fermentable carbohydrateson lipid metabolism in human subjectsremains to be elucidated. This is par-tially because the putative metabolic tar-gets of inulin, relatively well known anddescribed in animal models, are more dif-ficult to study in humans. Several factorsshould be taken into account in humanstudies, including dietary intakes of car-bohydrates vs. lipids in the backgrounddiet, duration of treatment, and serum lipidcomposition at the beginning of the treat-ment as important variables that may in-fluence outcome (Jackson et al., 1999).
Further consideration and research mustconcentrate on (1) defining the mechanismsunderlying the putative cholesterol-lower-ing effects of probiotics, (2) the problems ofsurvivability in the large bowel, (3) the pos-sible role of prebiotics in maintaining ad-equate survival and colonisation of appro-priate cholesterol-lowering species, and (4)identifying the actual effects on human post-prandial triglycerides of prebiotics.
REFERENCES
Agerbaek, M., Gerdes, L.U., and Richelsen, B.1995. Hypocholesterolemic effect of a newfermented milk product in healthy middle-aged men. Eur. J. Clin. Nutrit. 49: 346–352.
Akalin, A.S., Gönç, S., and Düzel, S. 1997.Influence of yoghurt and acidophilus yo-ghurt on serum cholesterol levels in mice.J. Dairy Sci. 80(11): 2721–2725.
Andersson, H., Bosaeus, I., Ellegard, L., Grahn,E., Tidehag, P., Hallmans, G., Holm, S.,and Sandberg, A.S. 1995. Effects of low-fat milk and fermented low-fat milk oncholesterol absorption and excretion inileostomy subjects. Eur. J. Clin. Nutrit.49: 274–281.
Cri
tical
Rev
iew
s in
Bio
chem
istr
y an
d M
olec
ular
Bio
logy
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f C
alga
ry o
n 09
/10/
13Fo
r pe
rson
al u
se o
nly.
278
Andersson, H.B.I. 1993. Sterol balance studiesin man. A critical review. Eur. J. Clin.Nutrit. 47: 153–159.
Bertolami, M.C., Faludi, A.A., and Batlouni, M.1999. Evaluation of the effects of a newfermented milk product (Gaio) on primaryhypercholesterolemia. Eur. J. Clin. Nutrit.53: 97–101.
Brown, L., Rosner, B., Willett, W.W., and Sacks,F.M. 1999. Cholesterol-lowering effectsof dietary fiber: a meta-analysis. Am. J.Clin. Nutrit. 69: 30–42.
Canzi, E., Brighenti, F., Casiraghi, M.C., DelPuppo, E., and Ferrari, A. 1995. Prolongedconsumption of inulin in ready-to-eatbreakfast cereals: effects on intestinal eco-system, bowel habits and lipid metabo-lism. In: Proceedings of COST Action 92workshop-Dietary fibre and fermentationin the colon. pp. 280–284, Finland.
Causey, J.L., Feirtag, J.M., Gallaher, D.D.,Tungland, B.C., and Slavin, J.L. 2000.Effects of dietary inulin on serum lipids,blood glucose and the gastrointestinal en-vironment in hypercholesterolemic men.Nutrit. Res. 20(2): 191–201.
Chen, W-J. L., Anderson, J.W., and Jennings,D. 1984. Propionate may mediate thehypocholesterolemic effects of certainsoluble plant fibers in cholesterol-fed rats.Proceed. Soc. Exp. Biol. Med. 175: 215–218.
Daubioul, C.A., Taper, H.S., De Wispelaere,L.D., and Delzenne, N.M. 2000. Dietaryoligofructose lessens hepatic steatosis, butdoes not prevent hypertriglyceridemia inobese Zucker rats. J. Nutrit. 130: 1314–1319.
Davidson, M.H., Maki, K.C., Synecki, C., Torri,S.A., and Drennan, K.B. 1998. Effects ofdietary inulin on serum lipids in men andwomen with hypercholesterolemia. Nutrit.Res. 18: 503–517.
Delzenne, N.M., and Kok, N.N. 1999. Biochemicalbasis of oligofructose-induced hypolipidemia
in animal models. J. Nutrit. 129 (suppl):1467S-1470S.
Delzenne, N.M., Kok, N., Fiordaliso, M.F., Deboyser,D.M., Goethals, F.M., and Roberfroid, M.B.1993. Dietary fructooligosaccharides modifylipid metabolism in rats. Am. J. Clin. Nutrit.57(suppl): 820S.
De Rodas, B.Z., Gilliland, S.E., and Maxweell,C.V. 1996. Hypocholesterolemic action ofLactobacillus acidophilus ATCC 43121 andcalcium in swine with hypocholesterolemiainduced by diet. J. Dairy Sci. 79: 2121–2128.
De Roos, N.M., Schouten, G., and Katan, M.B.1999. Yoghurt enriched with Lactobacil-lus acidophilus does not lower blood lip-ids in healthy men and women with nor-mal to borderline high serum cholesterollevels. Eur. J. Clin. Nutrit. 53: 277–280.
De Smet, I., De Boever, P., and Verstraete, W.1998. Cholesterol lowering in pigs throughenhanced bacterial bile salt hydrolase ac-tivity. Br. J. Nutrit. 79: 185–194.
De Smet, I., Van Hoorde, L., De Saeyer, N.,Vande Woestyne, M., and Verstraete, W.1994. In vitro study of bile salt hydrolase(BSH) activity of BSH isogenic Lactoba-cillus plantarum 80 strains and estimationof cholesterol lowering through enhancedBSH activity. Microbial Ecol. Health Dis.7: 315–329.
Ellegard, L., Andersson, H., and Bosaeus, I.1997. Inulin and oligofructose do not in-fluence the absorption of cholesterol, orthe excretion of cholesterol, Ca, Mg, Zn,Fe, or bile acids but increases energy ex-cretion in ileostomy subjects. Eur. J. Clin.Nutrit. 51: 1–5.
Fernandez, M.L. 1995. Distinct mechanisms ofplasma LDL lowering by dietary fibber inthe guinea pig: specific effects of pectin,guar gum, and psyllium. J. Lipid Res. 36:2394–2404.
Fiordaliso, M., Kok, N., Desaher, J.P., Goethals,F., Deboyser, D., Roberfroid, M., and
Cri
tical
Rev
iew
s in
Bio
chem
istr
y an
d M
olec
ular
Bio
logy
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f C
alga
ry o
n 09
/10/
13Fo
r pe
rson
al u
se o
nly.
279
Delzenne, N. 1995. Dietary oligofructoselowers triglycerides, phospholipids andcholesterol in serum and very low densitylipoproteins of rats. Lipids 30: 163–167.
Fuller, R. and Gibson, G. 1998. Probiotics andprebiotics. In : Encyclopaedia of FoodMicrobiology. pp. 1633–1639, AcademicPress.
Gibson, G. 1999. Dietary modulation of thehuman gut microflora using the prebioticsoligofructose and inulin. J. Nutrit. 129:1438S-1441S.
Gibson, G.R., and Roberfroid, M.B. 1995. Di-etary modulation of the human colonicmicrobiota: introducing the concept ofprebiotics. J.Nutrit. 125: 1401–1412.
Gilliland, S.E. and Walker, D.K. 1990. Factorsto consider when selecting a culture ofLactobacillus acidophilus as a dietary ad-junct to produce a hypocholesterolemiceffect in humans. J. Dairy Sci. 73: 905–911.
Gilliland, S.E., Nelson, C.R., and Maxwell, C.1985. Assimilation of cholesterol by Lac-tobacillus acidophilus. Appl. Environ.Microbiol. 49: 377–381.
Groot, P.H.E., Van Stiphout, W.A.H.J., Krauss,X.H., Jansen, H., Van Tol, A., VanRamshort, E., Chin-On, S., Hofman, A.,Cresswell, S.R., and Havekes, L. 1991.Postprandial lipoprotein metabolism innormolipidemic men with and withoutcoronary artery disease. Arterioscler.Thromb. 11: 653–662.
Harrison, V.C. and Peat, G. 1975. Serum cho-lesterol and bowel flora in the newborn.Am. J. Clin. Nutrit. 28: 1351–1355.
Hepner, G., Fried, R., Jeor, S. St., Fusetti, L.,and Morin, R. 1979. Hypocholesterolemiceffect of yoghurt and milk. Am. J. Clin.Nutrit. 32: 19–24.
Hokanson, J.E., and Austin, M.A. 1996. Plasmatriglyceride level is a risk factor for car-diovascular disease independent of
high-density lipoprotein cholesterol level,a meta-analysis of population-based pro-spective studies. J. Cardiovascul. Risk 3:213–219.
Howard, A.N. and Marks, J. 1977. Hypo-cholesterolemic effect of milk. Lancet2: 957.
Hylemon, P.B. and Glass, T.L. 1983. Biotransfor-mation of bile acids and cholesterol by theintestinal microflora. In : Human IntestinalMicroflora in Health and Disease.David J.Hentges, Eds., Academic Press, New York,pp.189–239.
Jackson, K.G., Taylor, G.R.J., Clohessy, A.M.,and Williams, C.M. 1999. The effect ofthe daily intake of inulin on fasting lipid,insulin and glucose concentrations inmiddle-aged men and women. Br. J. Nutrit.82: 23–30.
Kikuchi-Hayakawa, H., Shibahara-Sone, H.,Osada, K., Onodera-Masuoka, N., Ishikawa,F., and Watanuki, M. 2000. Lower plasmatriglyceride level in Syrian hamsters fed onskim milk fermented with Lactobacilluscasei strain Shirota. Biosci. Biotechnol.Biochem. 64(3): 466–475.
Klaver, F.A.M., and van der Meer, R. 1993. Theassumed assimilation of cholesterol bylactobacilli and Bifidobacterium bifidumis due to their bile salt-deconjugating ac-tivity. Appl. Environ. Microbiol. 59: 1120–1124.
Kok, N., Roberfroid, M., Robert, A., andDelzenne, N. 1996. Involvement of lipo-genesis in the lower VLDL secretion in-duced by oligofructose in rats. Br. J. Nutrit.76: 881–890.
Larsen, L.A., Raben, A., Haulrik, N., Hansen,A.S., Manders, M., and Astrop, A. 2000.Effect of 8 week intake of probiotic milkproducts on risk factors for cardiovasculardiseases. Eur. J. Clin. Nutrit. 54: 288–297.
Levine, G.N., Keaney, J.F., and Vita, J.A. 1995.Cholesterol reduction in cardiovascular dis-
Cri
tical
Rev
iew
s in
Bio
chem
istr
y an
d M
olec
ular
Bio
logy
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f C
alga
ry o
n 09
/10/
13Fo
r pe
rson
al u
se o
nly.
280
ease. Clinical benefits and possible mecha-nisms. N. Engl. J. Med. 332: 512–521.
Levrat, M.A., Favier, M.L., Moundras, C., Remesy,C., Demigne, C., and Morand, C. 1994. Roleof dietary propionic acid and bile acid excre-tion in the hypocholesterolemic effects ofoligosaccharides in rats. J. Nutrit. 124: 531–538.
Levrat, M.A., Remesy, C., and Demigne, C.1991. High propionic acid fermentationsand mineral accumulation in the cecum ofrats adapted to different levels of inulin. J.Nutrit. 121: 1730–1737.
Lipid Research Clinics Program 1984. The lipidresearch clinics coronary primary preven-tion trial results. I. Reduction in incidenceof coronary heart disease. J. Am. Med.Assoc. 251: 351–363.
Luo, J., Rizkalla, S.W., Alamowitch, C.,Boussairi, A., Blayo, A., Barry, J-L,Laffitte, A., Guyon, F., Bornet, F.R.J., andSlama, G. 1996. Chronic consumption ofshort-chain fatty fructooligosaccharides byhealthy decreased basal hepatic glucoseproduction but had no effect on insulin-stimulated glucose metabolism. Am. J.Clin. Nutrit. 63: 939–945.
Mann, G.V. 1974. Studies of a surfactant andcholesterolemia in the Maasai. Am. J. Clin.Nutrit. 27: 464–469.
Meei-YN Lin and Tseng-Wei Chen 2000. Re-duction of cholesterol by Lactobacillusacidophilus in culture broth. J. Food DrugAnal. 8(2): 97–102.
Morgan, L.M. 1996. The metabolic role of GIP:physiology and pathology. Biochem. Soc.Transac. 24: 585–591.
Noh, D.O., Kim, S.H., and Gilliland, S.E. 1997.Incorporation of cholesterol into the cellu-lar membrane of Lactobacillus acidophilusATCC 43121. J. Dairy Sci. 80: 3107–3113.
Patsch, W. 1992. Relation of triglyceride me-tabolism and coronary artery disease.Arterioscler. Thromb. 12: 1336–1345.
Pedersen, A., Sandstrom, B., and Van Amelsvoort,J.M.M. 1997. The effect of ingestion of inu-lin on blood lipids and gastrointestinal symp-toms in healthy females. Br. J. Nutrit. 78:215–222.
Rasic, J.L., Vujcic, I.F., Skrinjar, M., and Vulic,M. 1992. Assimilation of cholesterol bysome cultures of lactic acid bacteria andbifidobacteria. Biotechnol Lett 14: 39–44.
Ratcliffe, H.L. and Luginbuhl, H. 1971. Domes-tic pig: a model for experimental athero-sclerosis. Atherosclerosis 13: 133.
Richelsen, B., Kristensen, K., and Pedersen, S.B.1996. Long-term (6 months) effect of anew fermented milk product on the levelof plasma lipoproteins — a placebo-con-trolled and double blind study. Eur. J. Clin.Nutrit. 50: 811–815.
Roberfroid, M.B. 1996. Functional effects offood components and the gastrointestinalsystem: chicory fructooligosaccharides.Nutrit. Rev. 54: S38–S42.
Rossi, E.A., Vendramini, R.C., Carlos, I.Z., Pei,Y.C., and de Valdez, G.F. 1999. Develop-ment of a novel fermented soymilk prod-uct with probiotic properties. Eur. FoodRes. Technol. 209: 305–307.
Rossouw, J.E., Burger, E.M., Van der Vyer, P.,and Ferreira, J.J. 1981. The effect of skimmilk, yoghurt and full cream milk on hu-man serum lipids. Am. J. Clin. Nutrit. 34:351–356.
Sanders M.E. 2000. Considerations for use ofprobiotic bacteria to modulate humanhealth. J. Nutrit. 130: 384S-390S.
Schaafsma, G., Meuling, W.J.A., van Dokkum,W., and Bouley, C. 1998. Effects of a milkproduct, fermented by Lactobacillus aci-dophilus and with fructo-oligosaccharidesadded, on blood lipids in male volunteers.Eur. J. Clin. Nutrit. 52: 436–440.
Schaarmann, G., Schneider, J., Zorn, A., Vilser,C., and Jahreis, G. 2001. Influence ofprobiotic yoghurt on serum lipids in
Cri
tical
Rev
iew
s in
Bio
chem
istr
y an
d M
olec
ular
Bio
logy
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f C
alga
ry o
n 09
/10/
13Fo
r pe
rson
al u
se o
nly.
281
women. Am. J. Clin. Nutrit. 73(Suppl.):496S.
Sessions, V.A., Lovegrove, J.A., Dean, T.S., Wil-liams, C.M., Sanders, T.A.B., Macdonald,I.A., and Salter, A.M. 1998. The effect of anew fermented milk product on plasma cho-lesterol and apolipoprotein B concentrationsin middle-aged men and women. In : Func-tional Foods the consumer, the products andthe evidence. Michele J. Sadler Eds., RoyalSociety of Chemistry, UK, pp. 15–19.
Shaper, A.G., Jones, K.W., Jones, M., andKyobe, J. 1963. Serum lipids in three no-madic tribes of northern Kenya. AmericanJ. Clin. Nutrit. 13: 135–146.
St-Onge, M.P., Farnworth, E.R., and Jones,P.J.H. 2000. Consumption of fermentedand nonfermented dairy products: effectson cholesterol concentrations and metabo-lism. Am. J. Clin. Nutrit. 71: 674–681.
Tahri, K., Crociani, J., Ballongue, J., andSchneider, F. 1995. Effects of three strainsof bifidobacteria on cholesterol. Lett. Appl.Microbiol. 21: 149–151.
Tahri, K., Grill, J.P., and Schneider, F. 1996.Bifidobacteria strain behaviour towardcholesterol: coprecipitation with bile saltsand assimilation. Curr. Microbiol. 33: 187–193.
Taranto, M.P., Medici, M., Perdigon, G., RuizHolgado, A.P., and Valdez, G.F. 2000.Effect of Lactobacillus reuteri on the pre-vention of hypercholesterolemia in mice.J. Dairy Sci. 83: 401–403.
Taylor, G.R.J., and Williams, C.M. 1998. Ef-fects of probiotics and prebiotics on bloodlipids. Br. J. Nutrit. 80 suppl. 2: S225–S230.
Trautwein, E.A., Rieckhoff, D., and Erbersdobler,H.F. 1998. Dietary inulin lowers plasmacholesterol and triacylglycerol and altersbiliary bile acid profile in hamsters. J. Nutrit.128: 1937–1943.
Usman and Hosono, A. 1999. Bile tolerance,taurocholate deconjugation, and bindingof cholesterol by Lactobacillus gasseristrains. J. Dairy Sci. 82: 243–248.
Usman and Hosono, A. 2000. Effect of admin-istration of Lactobacillus gasseri on se-rum lipids and fecal steroids in hypercho-lesterolemic rats. J. Dairy Sci. 83 (8):1705–1711.
Vanhoof, K. and De Schrijver, R. 1995. Effectof unprocessed and baked inulin on lipidmetabolism in normo- and hypercholes-terolemic rats. Nutrit. Res. 15: 1637–1646.
Walker, D.K. and Gilliland, S.E. 1993. Rela-tionships among bile tolerance, bile saltdeconjugation and assimilation of choles-terol by Lactobacillus acidophilus. J. DairySci. 76: 956–961.
Williams, C.M. 1999. Effects of inulin on lipidparameters in humans. J. Nutrit. 129(suppl): 1471S–1473S.
Wolever, T.M.S., Fernandes, J., and VeketeshwerRao, A. 1996. Serum acetate: propionate ratiois related to serum cholesterol in men but notin women. J. Nutrit. 126: 2790–2797.
Yamashita, K., Kawai, K., and Itakura, M. 1984.Effects of fructo-oligosaccharides on bloodglucose and serum lipids in diabetic sub-jects. Nutrit. Res. 4: 961–966.
Young, J. 1998. European market developments inprebiotic- and probiotic-containing foodstuffs.Br. J. Nutrit. 80 (suppl. 2): S231–S233.
Cri
tical
Rev
iew
s in
Bio
chem
istr
y an
d M
olec
ular
Bio
logy
Dow
nloa
ded
from
info
rmah
ealth
care
.com
by
Uni
vers
ity o
f C
alga
ry o
n 09
/10/
13Fo
r pe
rson
al u
se o
nly.