We report successful vitamin A supplementation by inhalation of retinyl palmitate in a placebo-controlled pilot study in twenty-five preschool children (2–5 years of age) in the rural district of Gondar, Ethiopia. Preschool children (n 161) were randomly selected from 220 households. Out of this cohort, twenty-five children were randomly assigned to each of two treatment groups: one receiving retinyl palmitate by inhalation of two puffs of an aerosol containing 1 mg (3000 IU) per delivery to give a total of 2 mg (6000 IU); and the other receiving an aerosol without retinyl palmitate. Both treatments were administered every 2 weeks for 3 months. Serum retinol and retinol-binding protein concentrations in the vitamin A-treated group were 0·68 (sd 0·31) μmol/l and 59·4 (sd 24·2) mg/l before and 1·43 (sd 0·46) μmol/l (P < 0·01) and 97·3 (sd 31·2) mg/l (P < 0·05) 3 months after supplementation with retinyl palmitate, suggesting that this novel method of delivery of retinyl palmitate by inhalation is effective in improving vitamin A status.

Twenty-nine women and fifteen men from an area of low Se intake (South Island of New Zealand) consumed 100 μg stable 74Se, as selenate given in water after an overnight fast, and blood was collected for 3 weeks. They were then divided into five groups and supplemented with 0, 10, 20, 30 and 40 μg Se/d (as selenomethionine) for 5 months. After 5 months, they received a second dose of 74Se identical to the first. Supplementation significantly altered retention of 74Se in the plasma, but not in the erythrocytes or platelets. Subjects receiving the placebo retained the greatest amount, and subjects receiving 30 μg supplemental Se/d retained the least 74Se. Supplementation resulted in relatively more isotope being retained in a medium molecular mass protein considered to be albumin, and relatively less in another fraction considered to be selenoprotein P. The lack of many observed changes in retention of stable Se, and the shift in retention among the plasma proteins, suggests that supplemental Se was not being used to replete critical pools of Se, probably because of adaptation to low Se intake.

The fatty acid compositions of white adipose tissue, colostrum and mature milk triacylglycerols from Mauritian (n 13) and French (n 15) women were analysed and compared in order to highlight cultural differences in dietary intakes and their influence on milk fatty acid composition. Erythrocyte phosphatidylethanolamine and phosphatidylcholine fatty acid compositions were also investigated in their term infants, breast-fed over a period of 6 weeks. Fatty acid composition (g/100 g) of all samples was determined by GLC and anthropometric measurements were assessed in the two populations at birth and on day 42. Comparisons of white adipose tissue fatty acid compositions demonstrated lower levels of saturated (23·64 (se 1·54) v. 29·75 (se 0·67), P < 0·01) and monounsaturated (39·44 (se 1·27) v. 54·84 (se 0·75), P < 0·001) fatty acids and higher levels of polyunsaturated fatty acids (n−6 series: 32·47 (se 1·31) v. 14·32 (se 0·47), P < 0·001 and n−3 series: 2·87 (se 0·49) v. 0·80 (se 0·07), P < 0·01) in Mauritian than in French samples respectively. Accordingly, milk fat of the Mauritian women contained higher levels of parent essential fatty acids and their longer-chain derivatives than did milk fat from French women. Higher levels of parent essential fatty acids but lower levels of long-chain polyunsaturated fatty acids were found in erythrocyte phospholipids of Mauritian infants compared with French infants. Infants' erythrocyte arachidonate and docosahexaenoate contents did not correlate with any anthropometric variables at birth or at day 42, neither did they correlate with anthropometric variation over the study period. Our results suggest the lack of a simple relationship between the amount of long-chain polyunsaturated fatty acids in human milk and their accretion in the erythrocyte phospholipids of breast-fed infants when provided concomitantly with high levels of both linoleic and α-linolenic acids in ratios which fall within recommended ranges.

An elevated plasma total homocysteine (tHcy) concentration is a risk factor for cardiovascular disease and for having offspring with a neural-tube defect. Folate is a methyl donor in the remethylation of homocysteine into methionine. Although folic acid supplementation decreases tHcy concentrations, effects of folic acid supplementation on plasma methionine concentrations are unclear. There is also concern that folic acid supplementation negatively affects vitamin B12 status. We studied effects of low-dose folic acid supplementation on methionine and vitamin B12 concentrations in plasma. We also investigated whether baseline plasma methionine and tHcy concentrations correlated with the baseline folate and vitamin B12 status. For a period of 4 weeks, 144 young women received either 500 μg folic acid each day, or 500 μg folic acid and placebo tablets on alternate days, or a placebo tablet each day. Plasma methionine, tHcy and plasma vitamin B12 concentrations were measured at start and end of the intervention period. Folic acid supplementation had no effect on plasma methionine or plasma vitamin B12 concentrations although it significantly decreased tHcy concentrations. Plasma methionine concentrations showed no correlation with either tHcy concentrations (Spearman rs - 0·01, P = 0·89), or any of the blood vitamin variables at baseline. Baseline tHcy concentrations showed a slight inverse correlation with baseline concentrations of plasma vitamin B12 (rs - 0·25, P < 0·001), plasma folate (rs - 0·24, P < 0·01) and erythrocyte folate (rs - 0·19, P < 0·05). In conclusion, low-dose folic acid supplementation did not influence plasma methionine or plasma vitamin B12 concentrations. Furthermore, no correlation between plasma methionine concentrations and the blood folate and vitamin B12 status was shown.

Total substitution of soyabean protein for animal protein in the diet has been repeatedly shown to lower plasma cholesterol levels in hypercholesterolaemic individuals. A new, highly palatable, high-protein soya drink may allow replacement of a significant percentage of animal protein in the diet. The soya drink was given, within a crossover design v. a cows' milk preparation of similar composition and taste, to twenty-one severely hypercholesterolaemic patients (mean baseline plasma cholesterol 8·74 mmol/l) with a history of resistance to or intolerance of statin treatment. Each dietary supplement was given for 4 weeks, with a 4-week interval between treatments, Plasma lipid levels were monitored every 2 weeks during each dietary sequence. The concomitant dietary treatment, which had been followed for a long time by all patients, was carefully monitored throughout the study. The soya supplementation reduced plasma total cholesterol level by 6·5%, when given first, and by 7·4% when given after cows' milk. When given first, cows' milk resulted in a small, non-significant reduction of plasma cholesterol level (-3·9%), and when given after soya, it changed total plasma cholesterol to a minimal extent (-1·6%). Changes in total and LDL-cholesterol levels after 2 and 4 weeks of soya v. cows' milk treatment were, thus, respectively -6·1, -7·0 and -6·2, -7·8% (both P < 0·05). These first data from a double-blind study confirm a significant cholesterol-lowering effect of soyabean protein, even when only partly replacing animal protein in the diet, in individuals with extreme plasma cholesterol elevations.

Consumption of unrefined whole flour is thought to affect mineral bioavailability because it contains high levels of fibre and phytic acid. The present experiment was designed to study the absorption of minerals from diets based on wholewheat flour and white wheat flour in rats. Two groups of male Wistar rats were fed on the diets for 3 weeks and absorption and tissue retention of minerals were studied. The rats fed on the wholewheat flour diet had significantly greater food intake, weight gain, faecal excretion and intestinal fermentation than those fed on the white flour diet. Mineral intakes, except for Ca, were significantly greater in rats fed on the wholewheat flour diet (4-fold for Mg, 2-fold for Fe and Zn). A significant rise in the apparent absorption of Fe (%) and a significant decrease in the apparent absorption of Zn (%) were observed. The amounts of minerals absorbed (mg/d) were significantly enhanced (excepted for Ca) with the wholewheat flour diet. Moreover, plasma and tibia levels of Mg and plasma, liver and tibia levels of Fe were significantly increased in rats fed on the wholewheat flour diet compared with those fed on the white flour diet. In conclusion, wholewheat flour, rich in phytic acid and minerals, did not have a negative effect on mineral absorption, but rather improved the bioavailability of some minerals. Human studies are needed to confirm these rat results before extrapolation to human nutrition.

In a 9-week study seventy-six healthy adult volunteers with an average age of 44 (sd 11) years, with baseline plasma total cholesterol levels below 8 mmol/l, received in a balanced, double-blind, crossover design, a total of three different table spreads for personal use. Two spreads were fortified either with free (non-esterified) vegetable-oil sterols, mainly from soyabean oil (31 g sterol equivalents/kg; 0·8 g/d) or sheanut-oil sterols (133 g sterol equivalents/kg; 3·3 g/d). One spread was not fortified (control). Average intake of spread was 25 g/d for 3 weeks. None of the spreads induced changes in blood clinical chemistry or haematology. Plasma total- and LDL-cholesterol concentrations were statistically significantly reduced by 3·8% and 6% (both 0·19 mmol/l) respectively, for the spread enriched with free soyabean-oil sterols compared with the control spread. The spread enriched with sheanut-oil sterols did not lower plasma total- and LDL-cholesterol levels. None of the plant-sterol-enriched spreads affected plasma HDL-cholesterol concentrations. Plasma-lipid-standardized concentrations of α- plus β-carotene were not statistically significantly affected by the soyabean-oil sterol spread in contrast to lipid-standardized plasma lycopene levels which showed a statistically significant decrease (9·5%). These findings indicate that a daily intake of free soyabean-oil sterols as low as 0·8 g added to a spread is effective in lowering blood total- and LDL-cholesterol levels with limited effects on blood carotenoid levels. The lowering in total- and LDL-cholesterol blood levels due to consumption of the vegetable-oil-sterol-enriched spread may be helpful in reducing the risk of CHD for the population.

In the present intervention study we compared the effects of α-linolenic acid with those of eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) on platelet aggregation in vitro and ex vivo in healthy non-institutionalized elderly subjects. We also compared the effects of α-linolenic acid on platelet aggregation in elderly subjects with those in younger volunteers. During a run-in period of 3 weeks all subjects (thirty-eight elderly (> 60 years) and twelve younger volunteers (< 35 years)) received a diet rich in oleic acid. For the next 6 weeks the elderly subjects received a diet rich in oleic acid (n 11), α-linolenic acid (n 14) (6·8 g/d) or EPA/DHA (n 13) (1·05 g EPA plus 0·55 g DHA). The younger subjects were given a diet rich in α-linolenic acid. The diets did not affect ADP- or collagen-induced platelet aggregation in vitro in either platelet-rich plasma or whole blood. The ex vivo platelet aggregation as measured with filtragometry was significantly decreased in the elderly group that received EPA/DHA compared with the α-linolenic acid (P = 0·006) and the oleic acid (P = 0·005) diet groups. Effects of α-linolenic acid were not age-dependent. Our results suggest that α-linolenic acid and EPA/DHA do not change in vitro platelet aggregation. Compared with oleic acid, EPA/DHA, but not α-linolenic acid, favourably affects ex vivo platelet aggregation in healthy elderly subjects.

Human milk oligosaccharides are known to play a role in protection against certain infectious diseases. Previous reports indicate that the content of human milk oligosaccharides varies widely among individuals at term but such information on preterm milk is lacking. After removal of the fat, protein and most of the lactose from non-pooled human milk samples, a total neutral oligosaccharide fraction was isolated by ion-exchange chromatography followed by gel filtration. A Dionex high-performance anion-exchange chromatography system equipped with a pulsed electrometric detector was then employed to measure the levels of ten neutral oligosaccharides in the individual milk samples. Twenty-three milk samples from thirteen mothers who delivered at a mean gestational age of 29·5 (sd 3·1) weeks were collected between days 0 and 33 of lactation, and compared with three samples of term milk from two mothers. The ranges of the total and individual levels of the ten neutral oligosaccharides in preterm milk were similar to those in term milk. Further, as previously described in term milk, preterm milk exhibited a quantitative individual variation. This variation was independent of the gestational age, day of lactation, and postconceptional age. In conclusion, levels of ten neutral oligosaccharides did not differ between preterm and term human milk.

Cognitive performance has been found to decline after exposure to stress, particularly in stress-prone subjects. The present study investigated whether a carbohydrate-rich, protein-poor (CR/PP) diet, which may enhance cerebral serotonin function in stress-prone subjects due to increases in the available tryptophan, improves the performance of stress-prone subjects after exposure to acute laboratory stress. Twenty-two high-stress-prone (HS) subjects and twenty-one low-stress-prone (LS) subjects aged between 19 and 26 years performed a memory scanning task after controllable and uncontrollable stress, following either a CR/PP diet or a protein-rich, carbohydrate-poor (PR/CP) isoenergetic diet. Uncontrollable stress reduced feelings of control (F(1,38) 9·30; P = 0·004), whereas pulse rate and skin conductance increased after both stress tasks (F(1,38) 78·34; P = 0·0005 and F(1,37) 83·16; P = 0·0004). Diet, stress-proneness and stress-controllability interacted (F(1,36) 9·46; P = 0·004) in such a way that performance in HS subjects was better with the CR/PP diet than with the PR/CP diet, but only after controllable stress. As the CR/PP diet has been found to increase the plasma tryptophan: large neutral amino acids ratio, indicating an increased availability of cerebral tryptophan and, thus, higher serotonin levels, it appears that there may be an increased availability of brain serotonin in HS subjects after controllable laboratory stress.

Obesity has been on the increase among people of the Arabian Gulf countries. Overweight and obesity among 18–29-year-old Kuwaiti men increased by 23·4 and 14·8 % respectively, between 1980 and 1993. The objective of the present study was to explore factors associated with overweight and obesity in a sample of 515 Kuwaiti college men studied in 1997. Weight and height were measured. The index of adiposity used was the BMI, which is the weight (kg) divided by the height (m) squared (kg/m2). The men were classified as overweight (BMI > 25 kg/m2) or obese (BMI > 30 kg/m2). The associated factors obtained through questionnaires included age, marital status, governorate, number of siblings, suffering from a chronic disease, subjects' parental obesity, education and occupation, number of major meals eaten, eating between meals, family income, number of servants, number of people living at home, exercising, last dental and physical check-up, dieting, year of study, highest desired degree after college, countries preferred for visiting, and socio-economic status. The results of the study revealed that 38·5 and 11·1% of the students were overweight and obese respectively. Factors that were found to be significantly associated with overweight and obesity among the men included age, marital status, last dental check-up, exercising, subjects' parental obesity, dieting and year of study. Logistic regression analysis of significant associated factors revealed that the same factors contributed to the development of overweight and obesity.

There is increasing evidence that the degree of postprandial lipaemia may be of importance in the development of atherosclerosis and IHD. Postprandial lipid, lipoprotein, glucose, insulin and non-esterified fatty acid (NEFA) concentrations were investigated in eleven healthy young males after randomized ingestion of meals containing rapeseed oil, sunflower oil or palm oil with or without a glucose drink. On six occasions each subject consumed consecutive meals (separated by 1·75 h) containing 70 g (15 g and 55 g respectively) of each oil. On one occasion with each oil 50 g glucose was taken with the first meal. One fasting and fifteen postprandial blood samples were taken over 9 h. There were no statistically significant differences in lipoprotein and apolipoprotein responses after rapeseed, sunflower and palm oils, whereas insulin responses were lower after sunflower oil than after rapeseed oil (ANOVA, P = 0·04). The NEFA and triacylglycerol concentrations at 1·5 h were reduced when 50 g glucose was taken with the first meal (ANOVA, P < 0·0001 and P < 0·05 respectively), regardless of meal fatty acid composition. In conclusion, the consumption of glucose with a mixed meal containing either rapeseed, sunflower or palm oil influenced the immediate triacylglycerol and NEFA responses compared with the same meal without glucose, whereas no significant effect on postprandial lipaemia after a subsequent meal was observed. The fatty acid composition of the meal did not significantly affect the lipid and lipoprotein responses, whereas an effect on insulin responses was observed.

The present study was carried out to examine the effect of the daily intake of 10 g inulin on fasting blood lipid, glucose and insulin levels in healthy middle-aged men and women with moderately raised total plasma cholesterol (TC) and triacylglycerol (TAG) levels. This study was a double-blind randomized placebo-controlled parallel study in which fifty-four middle-aged subjects received either inulin or placebo for a period of 8 weeks. Fasting blood samples were collected before the supplementation period (baseline samples 1 and 2, separated by 1 week) and at weeks 4 and 8, with a follow-up at week 12. Compared with baseline values, insulin concentrations were significantly lower at 4 weeks (P < 0·01) in the inulin group. There was a trend for TAG values, compared with baseline, to be lower in the inulin group at 8 weeks (P < 0·08) returning to baseline concentrations at week 12. On comparison of the inulin and placebo groups, the fasting TAG responses over the 8-week test period were shown to be significantly different (P < 0·05, repeated measures ANOVA), which was largely due to lower plasma TAG levels in the inulin group at week 8. The percentage change in TAG levels in the inulin group during the 8-week study was shown to correlate with the initial TAG level of the subjects (rs −0·499, P = 0·004). We therefore conclude that the daily addition of 10 g inulin to the diet significantly reduced fasting insulin concentrations during the 8-week test period and resulted in lower plasma TAG levels, particularly in subjects in whom fasting TAG levels were greater than 1·5 mmol/l. These data support findings from animal studies that fructans influence the formation and/or degradation of TAG-rich lipoprotein particles, and the insulin data are also consistent with recent studies showing attenuation of insulin levels in fructan-treated rats.

Few studies have examined the association between specific sources of protein and blood lipids in a national sample of adults. We examined this relationship in a sample of adults 20 years and older who participated in phase 1 (1988–91) of the Third National Health and Nutrition Examination Survey, a representative sample of the United States non-institutionalized population. After excluding those participants who reported having been told they had high blood cholesterol concentrations, the final sample size was 6228. Mean intakes of different sources of proteins, as a percentage of total protein, were compared in quartiles of blood lipids. Intakes were adjusted for age, sex and race. Additional adjustments were made for other dietary variables, recall day, BMI, smoking, and income. We observed a lower percentage meat, fish and poultry (MFP) protein intake, including a lower percentage of beef and pork protein, among persons in the lowest quartile of serum total cholesterol and apolipoprotein B (ApoB) concentrations than among persons in the higher quartiles. The percentage of plant protein intake was higher in the lowest quartile than in the highest quartile of serum cholesterol. We also observed a higher percentage of fruit protein intake with lower serum cholesterol and ApoB concentrations. We conclude that in this cross-sectional sample, consumption of MFP proteins was consistently higher among persons with higher cholesterol concentrations while consumption of plant proteins was consistently higher among persons with lower cholesterol concentrations. Our findings support the importance of assessing intake of specific protein sources, especially in studies that address dietary intake in relation to blood lipids.

The influence of Cu supplementation of the usual diet for 6 weeks on biochemical markers of bone turnover and on putative indices of Cu status was investigated in healthy adults (twelve male and twelve female) aged 22–46 years, who participated in a double-blind placebo-controlled repeated crossover study. The study consisted of three 6-week supplementation regimens of 3 mg CuSO4, 3 mg Cu–glycine chelate (CuGC), and 6 mg CuGC, each separated by placebo periods of equal length. During baseline and on the last day of each dietary period, fasting morning first-void urine and fasting blood serum, plasma and erythrocytes were collected. The habitual dietary Cu intakes in males and females were approximately 1·4 and 1·1 mg/d respectively. Females had significantly higher (50 %) plasma caeruloplasmin (Cp) protein concentrations than males at baseline. Cu supplementation had no effect on erythrocyte superoxide dismutase (SOD, EC 1.15.1.1) activity or plasma Cp protein (putative indices of Cu status) in the total group. Similarly, serum osteocalcin (a marker of bone formation), urinary creatinine (Cr) concentration, urinary pyridinoline : Cr or deoxypyridinoline : Cr excretion (markers of bone resorption) were unaffected in either the total group or in males and females separately, by any Cu supplementation regimen. It is concluded that Cu supplementation of the usual diet in healthy adult males and females had no effect on biochemical markers of bone formation or bone resorption over 6-week periods.

Only a few studies have been published on the postprandial effects of different fatty acids in obese subjects. Therefore, the present study investigated the effects of three test meals containing palm oil (PO), lard (LD), or puff-pastry margarine (PPM), all normal dietary ingredients, on postprandial lipid and hormone responses in normal-weight and obese young women. The study was performed as a randomized, crossover design. The fats differed in the content of palmitic acid, stearic acid, and trans monounsaturated fatty acids allowing a dietary comparison of different ‘solid’ fatty acids. The obese women had significantly higher fasting concentrations and postprandial responses of plasma total triacylglycerol (TAG), chylomicron-TAG, and insulin compared with the normal-weight women but there was no significant difference in the postprandial responses between the three test meals. The obese women had fasting concentrations of leptin four times greater than the normal-weight women. There were no postprandial changes in the concentrations of leptin. The fasting concentrations of HDL-cholesterol were significantly lower in the obese women than in the normal-weight women, whereas there was no significant difference between the two groups in the concentrations of total cholesterol or LDL-cholesterol. These results provide evidence that obese women have exaggerated lipid and hormone responses compared with normal-weight women but the different contents of saturated and trans monounsaturated fatty acids provided by PO, LD, and PPM have no effect in either group.

It has been suggested that milk fat, due to its content of saturated fatty acids, may have a thrombogenic effect. In the present study the fatty acid profile of milk fat was modified by changing the feeding regimens of cows and the effect on haemostatic variables of a diet containing the modified milk fat (M) was compared with that of a diet containing milk fat of typical Danish composition (D). In the modified fat 16% of the saturated fatty acid (C12–C16) content was replaced mainly by oleic acid. Eighteen subjects were fed on two strictly controlled isoenergetic diets containing 40% energy from total fat (30% energy from the test fats) for periods of 4 weeks in a study with a crossover design. Fasting samples were taken in the last week of each study period. Postprandial samples were taken on day 21, 3 h after lunch (n 18), and on the last day of the study 2, 4, 6 and 8 h after a fat load containing 1·2 g of one of the milk fats/kg body weight (n 8). After 4 weeks' dietary intervention fasting plasma factor VII coagulant (FVIIc) activity, tissue-type plasminogen activator (t-PA) activity, plasminogen activator inhibitor (PAI-1) antigen and β-thromboglobulin did not differ between diets M and D. Postprandially FVIIc and t-PA activities increased (P < 0·001) and PAI-1 antigen and PAI-1 activity decreased (P < 0·001) as compared with fasting values, regardless of diet. After the fat load, the postprandial increase in FVIIc was marginally lower after diet M than diet D (diet effect, P < 0·05). In conclusion, the modified milk fat obtained by the applied feeding strategy had virtually the same effects on haemostatic variables as conventional milk fat.

The aim of the present study was to quantify the incorporation of eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) into plasma lipids after oral administration of n-3 fatty acid ethyl esters, since little is known about the rate and pattern of incorporation into plasma lipid fractions. In addition, we aimed to obtain preliminary information regarding EPA half-life, which is needed to establish an optimal dosing schedule. Five healthy volunteers ingested two 8·5 g doses of n-3 fatty acid ethyl esters daily for 7 d, supplying 6·0 g EPA/d and 5·3 g DHA/d. The fatty acid compositions of plasma phospholipids (PL), cholesteryl esters (CE) and triacylglycerols (TAG) were determined during supplementation and during a washout period of 7 d. Half-lives of EPA and DHA were calculated. The proportion of EPA in PL showed a 15-fold increase after 7 d (P < 0·001), while DHA showed a smaller increase (P < 0·01). In CE, EPA also increased (P < 0·05), while DHA did not increase at all. Remarkably, incorporation of DHA into TAG was even higher than that of EPA. Half-life of EPA in PL ranged from 1·63 to 2·31 d (mean 1·97 (se 0·15) d), whereas mean half-life of EPA in CE was 3·27 (se 0·56) d. In three subjects, washout of EPA and DHA from TAG seemed to follow a bi-exponential pattern, with a short half-life (< 1 d) in the initial phase and a half-life of several days in the second phase. In conclusion, EPA ethyl esters are rapidly incorporated into plasma lipids, especially into PL. The relatively long half-life of EPA in plasma would permit a dosing schedule with intervals of ≥12 h in supplementation studies.

The purpose of the present study was to investigate the effects of Se restriction on the excretion of Se in men who had consumed high levels of this element during their entire lives. With the use of stable isotopes of Se as selenite, the excretion of methylated Se in urine was investigated in Chinese men (n 10) who had habitual chronic high intakes of this element. The relationship between either urine Se or trimethylselenonium (TMSe) to the estimated long-term Se intake was not linear over the entire range of intake, which was also true for the infusion of labelled selenite. A non-linear relationship was also found between urine TMSe and urine Se both for TMSe arising from catabolism of endogenous body Se and that from infused selenite. The data suggest a close precursor–product relationship of urine Se and its TMSe component based on the nearly identical specific activities for these two selenocompounds. Although dimethylselenide in breath was not measured in the present study, combining urinary TMSe with this breath test may be more useful in the assessment of long-term Se status.

A controlled study with eight healthy free-living subjects was carried out, in which energy intake was adjusted to the individual energy requirements. On administration of inulin, blood lipids, the faecal microflora, short-chain fatty acids and accompanying gastrointestinal symptoms were characterized in order to investigate the long-term effect of inulin. During the run-in phase (8 d), subjects received a typical Western diet providing 45 % energy as fat and 40 % energy as carbohydrate. Subsequently, the subjects consumed a fat-reduced diet which provided 30 % energy as fat and 55 % energy as carbohydrate for a period of 64 d using inulin as a fat replacer. The amounts of inulin consumed by the subjects (up to 34 g/d) were based on individual energy requirements with the aim to keep the diet isoenergetic with that used in the run-in period. To assess the effects of inulin administration, a control study (run-in and intervention) was carried out in which subjects consumed the same diet but devoid of inulin during the whole course of the study. To investigate the effect of inulin on faecal flora composition total bacteria and bifidobacteria in the faeces were enumerated by in situ hybridization with 16S rRNA targeted oligonucleotide probes. Inulin significantly increased bifidobacteria from 9·8 to 11·0 log10/g dry faeces and caused a moderate increase in gastrointestinal symptoms such as flatulence and bloatedness, whereas blood lipids and short-chain fatty acids remained essentially unaffected.