点击显示 收起
1 From INSERM U341, Department of Diabetes, Hôtel-Dieu Hospital, Paris.
2 Supported by a grant from Syndifrais. 3 Address reprint requests to G Slama, Department of Diabetes, Hôtel-Dieu Hospital, 1 Place du Parvis Notre-Dame, 75181 Paris 04, France. E-mail: gerard.slama{at}htd.ap-hop-paris.fr.
ABSTRACT
Background: Ingestion of fermented dairy products induces changes in the equilibrium and metabolism of the intestinal microflora and may thus have beneficial effects on the host.
Objective: We compared the effects of chronic consumption of yogurt with (fresh) or without (heated) live bacterial cultures (Lactobacillus bulgaricus and Streptococcus thermophilus) on plasma glucose, insulin, triacylglycerols, cholesterol, fatty acids, and short-chain fatty acids.
Design: Two groups of 12 healthy men with or without lactose malabsorption were selected with use of a breath-hydrogen test after a 30-g lactose load. Subjects were randomly assigned in a crossover design to 500 g/d of either fresh or heated yogurt for 2 periods of 15 d each, separated by a 15-d washout interval.
Results: Chronic consumption of fresh or heated yogurt had no detrimental effects on plasma glucose, insulin, or fatty acid areas under the curve in response to acute ingestion of 500 g yogurt in healthy men with or without lactose malabsorption. There were also no detectable changes in fasting plasma glucose, insulin, fatty acid, triacylglycerol, or cholesterol concentrations. In contrast, plasma butyrate was higher (P < 0.03) and plasma propionate tended to be higher (P = 0.059) in subjects without lactose malabsorption after fresh yogurt consumption than after heated yogurt consumption. There were no significant changes in plasma acetate. In subjects with lactose malabsorption, 15 d of fresh yogurt consumption also increased propionate production compared with values at baseline (P < 0.04). In the same group, the production of breath hydrogen was lower after fresh yogurt consumption than after heated yogurt consumption (P < 0.01).
Conclusions: In men with lactose malabsorption, chronic consumption of yogurt containing live bacterial cultures ameliorated the malabsorption, as evidenced by lower breath-hydrogen excretion, but increased propionate concentrations. In subjects without lactose malabsorption, such yogurt tended to increase propionate and increased butyrate.
Key Words: Yogurt • live bacterial cultures • heated yogurt • short-chain fatty acids • plasma glucose • plasma insulin • plasma lipids • diet • breath-hydrogen test • lactose malabsorption • men
INTRODUCTION
Recent claims of beneficial effects of fermented dairy products on serum cholesterol concentrations have attracted much attention (1–3). Several studies showed that chronic consumption of fermented dairy products results in reduced serum cholesterol and LDL-cholesterol concentrations in humans (1–3). Many bacterial strains used in yogurt production have been selected as probiotics in recent years (4).
Although beneficial effects have been shown, the ingestion of dairy products is associated with lactose malabsorption in certain persons. Lactose malabsorption is the only well-defined cause of milk intolerance in humans and is caused by lactase deficiency. It is now well established that lactose malabsorption can be attenuated by the use of yogurt containing live cultures because of the inherent ß-galactosidase (lactase) activity of the yogurt bacteria (5, 6). The undigested carbohydrate is generally fermented by colonic bacteria into hydrogen, carbon dioxide, and short-chain fatty acids (SCFAs: acetate, propionate, and butyrate) (7). However, undigested lactose is normally manifested by increased hydrogen excretion, whereas lactose digestion increases SCFA production (8). The SCFAs may have beneficial effects on human glucose and lipid metabolism. Acetate can reduce serum fatty acids (9), which are an important factor in lowering tissue glucose utilization and inducing insulin resistance (10). On the other hand, long-term dietary supplementation with propionate was shown to decrease blood glucose in rats (11) and in humans (12, 13). Butyrate is mostly used by colonocytes as an energy source (14). Considering the possible different consequences of yogurt consumption in persons with lactose malabsorption and in those without, yogurt might exert different effects on blood lipids in the 2 populations.
As far as we know, only one study of the effects of yogurt on the metabolism of subjects with or without lactose malabsorption has been reported (15). The authors showed that the acute ingestion of fresh yogurt increased postprandial plasma insulin and fatty acid concentrations in subjects with lactose malabsorption. Ingestion of heated yogurt, however, resulted in a rise in plasma insulin concentrations but had no significant effect on plasma fatty acids in the same subjects (15). In subjects without lactose malabsorption, a transient rise in insulin and drop in fatty acids followed the ingestion of heated or unheated yogurt (15). Therefore, the present study was designed to compare the effects of chronic consumption of a yogurt with or without viable bacteria (Lactobacillus bulgaricus and Streptococcus thermophilus) on plasma glucose, insulin, fatty acids, and SCFAs in healthy subjects with or without lactose malabsorption.
SUBJECTS AND METHODS
Subjects
Twenty-four healthy male volunteers aged 20–60 y participated in the study. The clinical and biological characteristics of these subjects are shown in Table 1. Twelve subjects had lactose malabsorption and 12 did not. Lactose malabsorption was identified on the basis of breath-hydrogen responses to an oral 30-g lactose challenge as shown in Figure 1. Subjects were identified as having lactose malabsorption when the mass fraction of breath-hydrogen excretion was >20 ppm. The area under the breath-hydrogen curve was 6465 ± 120 ppm•3 h for subjects with lactose malabsorption and 1455 ± 100 ppm•3 h for subjects without lactose malabsorption. The ethical committee of the Hôtel-Dieu Hospital approved the study and written, informed consent was obtained from each volunteer.
View this table:
TABLE 1. Clinical and anthropometric characteristics of the subjects at the beginning of the study1
FIGURE 1.. Breath-hydrogen responses to a 30-g lactose challenge during selection of subjects with () and without (•) lactose malabsorption. n = 12 per group. *Significantly different from subjects without lactose malabsorption, P < 0.001.
Experimental design
Subjects were randomly assigned in a crossover design to 500 g/d of either fresh yogurt containing live bacterial cultures or heated yogurt for 2 periods of 15 d each, separated by a 15-d washout interval. The fresh yogurt (Danone, Le Plessis Robinson, France) used in the study was nonfat and was produced according to French regulations. The heated yogurt was produced according to the same procedure as the fresh yogurt, followed by pasteurization to destroy the lactic bacteria. The macronutrient composition of the 2 yogurts was similar. Per 100 g, the 2 types of yogurt contained 193 kJ energy, 4.9 g protein, 6.4 g carbohydrate, and 0.1 g fat. The fresh yogurt contained 1 x 107 colony-forming units/g of L. bulgaricus and S. thermophilus, whereas the heated yogurt contained at most 102 colony-forming units/g of L. bulgaricus and S. thermophilus. Subjects were given a total of 500 g yogurt to consume daily. The total daily dose was provided as 4 cups of 125 g each that the subjects could consume over 3 meals at their convenience. At the beginning and the end of each dietary period, subjects were submitted to a 3-h kinetic study of plasma glucose, insulin, fatty acids, and SCFAs as well as breath-hydrogen excretion after they consumed the total daily dose of yogurt (500 g) as breakfast after an overnight fast.
Throughout the study, the subjects were asked to maintain their usual lifestyle habits and to consume a low-fiber diet as prescribed by a dietitian. At the beginning and end of each dietary period, each subject received individual counseling by a dietitian. A low-fiber diet was prescribed individually according to data obtained from dietary questionnaires to keep initial energy intakes and nutrient proportions constant throughout the study. To quantify macronutrient composition, subjects were asked to complete dietary questionnaires at the beginning of each dietary period and 7-d dietary records at the end of each dietary period. Household measuring cups or spoons and food pictures were used to quantify portion sizes of foods eaten. When each subject returned his records, the dietitian checked the contents of the records and clarified any ambiguous information with the subject. These records were analyzed by using the computer program PROFILE DOSSIER V3 (Audit Conseil en Informatique Médicale, Bourges, France). The dietary database of this program comprises 400 foods or groups of foods representative of the French diet. French food contents were obtained from the Table Ciqual (16).
Chemical assays
Plasma glucose was measured by the glucose oxidase method with a glucose analyzer (Beckman, Fullerton, CA). Plasma insulin was measured by radioimmunoassay (ERIA Diagnostics Pasteur, Marnes la Coquette, France). The antiserum used in the test had a cross-reactivity of 100% with human insulin and of 40% with proinsulin. Triacylglycerols were measured with Biomérieux kits (Marcy-l'Etoile, France). Total cholesterol was assayed with Labintest kits (Aix-en Provence, France). HDL cholesterol was measured with Boehringer Mannheim kits (Meylan, France). Fatty acids were measured with a Unipath kit (Dardilly, France). Plasma acetate, butyrate, and propionate were measured by automated head-space gas chromatography with a 30-m DB1 capillary column and a flame ionization detector (J&W Scientific, Folsom, CA). Acetate, butyrate, and propionate were esterified by the addition of acidified methanol, followed by centrifugation for 10 min at 3000 x g at 4°C. The supernate was heated for 30 min at 65°C in a turntable gas chromatograph. The hydrogen concentration in expired air was measured by gas chromatography (DP microlyzer; Quintron Instruments, Milwaukee).
Statistical methods
The analysis took into account the design of the experiment and the distributions of the variables. The effects of the 2 yogurts were analyzed after one load of yogurt (day 0), after 15 d of chronic intake (day 15), and as the difference (day 15 - day 0).
The effect of yogurt was analyzed separately in the groups of subjects with or without lactose malabsorption because of a treatment x group interaction. Then the validity of the crossover design was tested for each parameter by an analysis of covariance of baseline results of the second period with baseline results of the first period as the covariate and the treatment of the first period as the main factor.
If the crossover design was validated, the effects of the 2 yogurts for continuous variables with normal distributions were compared by a 2-period crossover analysis with use of SAS for WINDOWS (version 6.12; SAS Institute Inc, Cary, NC). In parallel, a multiple analysis of variance for repeated and paired measurements was performed by using the CSS statistical package (StatSoft, Tulsa, OK). The main factors considered in the analysis were the following: treatment (fresh or heated yogurt), time (baseline or 2 wk), and order of randomization (fresh yogurt first or heated yogurt first). In the presence of a group x treatment or group x order interaction, the 2 groups (those with or without lactose malabsorption) were analyzed separately.
If an effect of treatment was detected during the test of covariance, the crossover design was rejected and only the results of the first period were used for statistical analysis. The crossover design was rejected for the breath-hydrogen test and plasma butyrate values of subjects with lactose malabsorption. Thus, only the results of the first period were analyzed with use of a nonparametric test (Mann-Whitney U test). Results are expressed as means ± SEMs.
RESULTS
Diets and body weight
Total daily energy, lipid, and carbohydrate intakes were stable in the 2 groups of subjects throughout the study (Table 2). However, an apparent slight but significant increase in calcium intake was detected after 15 d of fresh yogurt consumption in both groups of subjects and after 15 d of heated yogurt consumption in subjects with lactose malabsorption. Additionally, in subjects with lactose malabsorption, protein intake also increased after consumption of the 2 types of yogurt. The low protein and calcium intake values at day 0 might have been the result of the method by which these values were obtained. Information obtained by use of 7-d dietary records, as at the end of each dietary period, is more accurate than information obtained by use of dietary questionnaires, as at the beginning of the dietary periods. Values obtained by use of dietary questionnaires are often underestimates.
View this table:
TABLE 2. Dietary macronutrient composition and calcium intake during consumption of fresh and heated yogurt1
There was a treatment x group interaction for all of the indexes studied. Thus, the 2 groups of subjects were analyzed separately.
Plasma glucose and insulin
Fasting plasma glucose and insulin concentrations were stable during the experimental periods and were not significantly different between the fresh and the heated yogurt periods. The areas under the curve of plasma glucose and insulin concentrations were not significantly different between the fresh and the heated yogurt periods, nor between the beginning and the end of each 15-d dietary period (Table 3).
View this table:
TABLE 3. Fasting plasma lipid concentrations, areas under the curve (AUCs) of plasma indexes, and breath-hydrogen excretion over 3 h after a load of fresh or heated yogurt at the beginning (day 0) and at the end (day 15) of each dietary period1
Serum lipids
Fasting serum triacylglycerol, cholesterol, HDL-cholesterol, and fatty acid concentrations were stable during the experimental periods and were not significantly different between the fresh and the heated yogurt periods (Table 3). Additionally, the areas under the curve of fatty acid concentrations were not significantly different between the fresh and the heated yogurt periods, nor between the beginning and the end of each 15-d dietary period.
Plasma short-chain fatty acids
Areas under the curve were calculated for plasma acetate, propionate, and butyrate. There were no significant effects of either fresh or heated yogurt on the areas under the curve of plasma acetate concentrations in either group of subjects.
In subjects without lactose malabsorption, the area under the curve of the plasma propionate concentration tended to be higher after 15 d of fresh yogurt consumption than after 15 d of heated yogurt consumption. Furthermore, this area also tended to increase after 15 d of fresh yogurt consumption compared with baseline (NS).
In subjects with lactose malabsorption, no significant differences in plasma propionate areas under the curve were found between fresh and heated yogurt consumption. However, within-treatment analysis showed a significant elevation of the plasma propionate area under the curve after 15 d of fresh yogurt consumption compared with baseline.
The effects of the 2 yogurts were compared separately because of a treatment x group interaction. However, we aimed to study propionate production between subjects with and without lactose malabsorption. Thus, subjects were compared only during the same treatment. There was no significant difference in the ability of subjects with or without lactose malabsorption to produce propionate after consumption of either heated or fresh yogurt. Similar results were found in the 2 groups of subjects after one load (day 0) as well as after chronic consumption (15 d).
In subjects without lactose malabsorption, the results for plasma butyrate concentrations were similar to those for propionate. The area under the curve of the plasma butyrate concentration was higher after 15 d of fresh yogurt consumption than after 15 d of heated yogurt consumption. Additionally, the difference in the plasma butyrate concentration area under the curve between the beginning and the end of the dietary period was greater during the fresh yogurt period than during the heated yogurt period (P < 0.03). In subjects with lactose malabsorption, there was no significant effect of either yogurt on the areas under the curve of plasma butyrate concentration.
Breath-hydrogen excretion
Lactose maldigestion was ameliorated after the ingestion of either fresh or heated yogurt (which contained the same amount of lactose) after only a single load. In subjects with lactose maldigestion, the area under the breath-hydrogen curve was 6465 ± 120 ppm•3 h after a load of 30 g lactose at the time the subjects were selected for the study and decreased to 2174 ± 648 ppm•3 h (P < 0.001) after one load of heated yogurt and to 2400 ± 602 ppm•3 h (P < 0.001) after one load of fresh yogurt (Table 3). In subjects without lactose malabsorption, there was no significant difference in breath-hydrogen excretion between chronic fresh yogurt and heated yogurt consumption.
In subjects with lactose malabsorption, analysis of covariance showed a treatment effect on baseline values. Thus, the crossover analysis was rejected and only the results of the first period were taken into consideration. Breath-hydrogen area under the curve at the end of the 15-d dietary period was significantly lower after the consumption of fresh yogurt (n = 7) than after the consumption of heated yogurt (n = 5). Breath-hydrogen area under the curve for subjects with lactose maldigestion tended to increase after heated yogurt consumption (day 15 compared with day 0: P < 0.06).
The effects of the 2 yogurts were compared in the 2 groups separately because of the presence of a treatment x group interaction. However, we aimed to study differences in breath-hydrogen production between subjects with and without lactose malabsorption. Thus, subjects were compared only during the same treatment. One load of fresh yogurt in subjects with lactose malabsorption resulted in low breath-hydrogen production, which was similar to that in subjects without lactose malabsorption. After 15 d of fresh yogurt consumption, hydrogen production tended to be higher (P < 0.09) in subjects with lactose malabsorption than in subjects without lactose malabsorption. After one load of heated yogurt, breath-hydrogen production tended to be higher in subjects with lactose malabsorption than in subjects without. After 15 d of heated yogurt consumption, breath-hydrogen production was significantly higher in subjects with lactose malabsorption (P < 0.016).
DISCUSSION
This study showed that chronic consumption of yogurt containing live bacteria not only ameliorated the lactose maldigestion of healthy subjects but also increased plasma butyrate and propionate concentrations in men with or without lactose maldigestion. Amelioration of lactose maldigestion by yogurt was shown previously in studies comparing the effects of yogurt ingestion with that of ingestion of heated yogurt, milk, or both (17–19). Lerebours et al (19) showed in 16 lactase-deficient subjects that yogurt enhances lactose digestion and that this beneficial effect is destroyed by pasteurization. Similarly, breath-hydrogen excretion was lower after yogurt ingestion than after milk ingestion in institutionalized elderly persons (18). This effect of yogurt was partially eliminated when milk that was fermented and then pasteurized was ingested. In another study, Shermak et al (17) showed that in children with lactose malabsorption, ingestion of yogurt containing active live cultures resulted in fewer symptoms and lower rates of rise in the breath-hydrogen curve than did ingestion of milk, whereas pasteurized yogurt gave intermediate results. Similar results were obtained by using lactase. Addition of lactase to milk resulted in lower breath-hydrogen excretion than did milk ingestion only (20, 21). Ingestion of high-lactase yogurt reduced breath hydrogen compared with conventional yogurt in lactase-deficient subjects (6).
In the study by Lerebours et al (19) of subjects with lactose malabsorption, results of breath-hydrogen tests after 8 d of ingestion of yogurt or yogurt that was fermented and then pasteurized were not significantly different from results obtained after only 24 h of ingestion. In the present study, which entailed a controlled, crossover design, 15 d of fresh yogurt ingestion by subjects with lactose malabsorption resulted in lower breath-hydrogen production than did 15 d of heated yogurt ingestion. In the same group, the heated yogurt tended to increase breath-hydrogen excretion (day 15 compared with day 0). Indeed, there was an adaptation to yogurt containing live cultures when used over a long period. On the other hand, ingestion of heated yogurt and the absence of live bacterial cultures tended to further worsen the lactose maldigestion. Consequently, the beneficial effect of yogurt containing live bacterial cultures on breath-hydrogen excretion was due to both an immediate effect of lactic acid bacteria and an adaptation of the subjects' intestinal environment to chronic yogurt ingestion.
Curiously, the production of SCFAs did not change in parallel with breath-hydrogen excretion. Theoretically, the alleviation of lactose maldigestion should lead to lower plasma SCFA concentrations, not higher. However, there is no argument in the literature that supports this hypothesis. Whereas the effects of yogurts with or without live bacteria on hydrogen excretion are well documented, there are no recent data on the effects of yogurts with or without live bacteria on plasma SCFA concentrations. However, some data were published recently on the relation between hydrogen production and plasma SCFA concentrations during intestinal fermentation of fiber-like polysaccharides and resistant starch (22–24). A dose of 10 g transgalactooligosaccharides/d for 3 wk decreased breath-hydrogen excretion and increased fecal concentrations of bifidobacteria in healthy men (25). Three weeks of a higher dose (15 g transgalactooligosaccharides), however, increased the concentration of breath hydrogen by 130% without any detectable changes in the number of bifidobacteria or SCFA and bile acid concentrations in feces in healthy subjects (22). On the other hand, during a randomized controlled study, 2 wk of consumption of different resistant starches from high-amylose cornstarch increased the mean fecal butyrate-to-SCFA ratio without altering breath hydrogen (23). After two 4-wk periods of amylomaize starch (high in resistant starch) compared with available cornstarch (low in resistant starch), increased fermentation was verified by elevated breath-hydrogen excretion (24). Fecal concentrations and daily excretion of SCFAs were not significantly different in the 2 periods. Therefore, fermentation of nondigestible resistant starch or oligosaccharides might result in many conditions leading to the production of either hydrogen or SCFAs that depend on the type and dose of the factor used as well as the duration of the dietary period. Furthermore, in the present study, hydrogen production tended to increase after one load of heated yogurt and increased further after chronic consumption in subjects with lactose malabsorption compared with those without lactose malabsorption in the absence of any significant changes in plasma propionate.
Thus, there is a dissociation between the pathways leading to hydrogen production and those resulting in SCFAs. The fermentation procedure in the colon includes different steps, each accomplished by different bacteria (7). The human colon is essentially oxygen free; saccharolytic bacterial species oxidize glucose via the Ebden-Meyerof pathway to pyruvate. From pyruvate, various bacteria produce different products, such as hydrogen, lactate, ethanol, and succinate. Other bacteria then ferment these intermediates to SCFAs (7). Any change in the balance between these different bacteria may result in a different ratio of hydrogen to SCFAs. In an in vitro continuous culture system inoculated with fresh samples of human feces and infused with lactose, supplementation with L. acidophilus resulted in a significant decrease in lactose concentration and a greater increase in acetate and propionate production than in the control group (26). Bifidobacterial supplementation also enhanced lactose digestion and increased acetate production (8). Thus, lactose maldigestion is manifested by increased hydrogen excretion, whereas alleviation of lactose maldigestion might be associated with increased plasma SCFAs as in the present study.
It was suggested that high plasma concentrations of SCFAs ameliorate plasma glucose and lipid concentrations. This concept was strengthened in part by a recent study in our laboratory in which 3 wk of feeding rats a diet rich in propionate decreased plasma glucose concentrations (11). Few studies tested this hypothesis in humans. Administration of 7.5 g sodium propionate/d for 7 wk lowered fasting plasma glucose and ameliorated glucose tolerance in a group of 10 healthy women (12). Moreover, Todesco et al (13) found that chronic intake of 9.9 g sodium propionate in bread for 1 wk decreased plasma glucose concentrations in response to a bread challenge containing 50 g carbohydrate. In the present study, the increase in plasma propionate and butyrate in subjects with or without lactose malabsorption induced by 15 d of yogurt ingestion did not result in any detectable beneficial effect on plasma glucose, insulin, or lipids. This might have been due either to insufficient amounts of propionate and butyrate produced or the need for longer periods to detect an effect.
In the literature, the reported effects of yogurt ingestion on plasma glucose and lipids are inconsistent. In subjects with lactose malabsorption, Dewit et al (15) showed an increase in plasma insulin and fatty acid concentrations after a single load of fresh yogurt. In subjects without lactose malabsorption, a transient rise in insulin and a drop in fatty acids was found. A reduction in serum cholesterol and LDL cholesterol after chronic consumption of fermented dairy products was observed in some studies (1–3) but not others (27). Effects of yogurt on plasma lipids seem to be related in part to the bacterial strains used in milk fermentation (2).
In conclusion, the chronic consumption of yogurt containing live bacterial cultures ameliorated the maldigestion in men with lactose malabsorption. In men with and without lactose malabsorption, fresh yogurt consumption increased butyrate and propionate production, which may ameliorate glucose and lipid metabolism in the long term. Thus, chronic yogurt consumption may be of benefit to both subjects with lactose malabsorption and those without.
ACKNOWLEDGMENTS
We thank B Guy Grand (Nutrition Department, Hôtel-Dieu Hospital) for the opportunity to perform measurements in the Nutrition Department's laboratory.
REFERENCES