Peripheral vascular disease (PVD) is a manifestation of systemic atherosclerosis in the lower limbs, and PVD patients have a 3- to 5-fold increased risk of cardiovascular mortality compared with age-matched controls. Nevertheless, recent reports show how PVD patients are undertreated with regard to CVD risk-factor reduction and the use of lipid-lowering or antiplatelet drugs. There is appreciable evidence that demonstrates the beneficial effects of certain nutrients and dietary habits in the prevention of CVD, but there has been little attention paid to the role of nutrients in PVD. The purpose of the present review is to provide an overview of our understanding of how foods could possibly benefit PVD. In the last few decades, several nutrients have arisen as potentially health-promoting in PVD. While nutritional interventions in PVD show positive clinical effects for fish oil, carnitine or vitamin E, others such as olive oil or vitamin C seem to interact only at a biochemical level by decreasing risk factors. Moreover, only epidemiological associations exist for the potential role of fibre, folates or vitamin B6 in this disease. In all cases, the limited data available provide no clear-cut evidence in favour of the clinical benefit of nutritional interventions aimed at reducing risk factors and ameliorating symptoms in PVD patients. No practical recommendations can be given at this stage, and further studies are clearly needed.

The objective of this investigation was to determine the influence of flaxseed oil on responses representative of primary humoral and cell-mediated adaptive immune competence in immunosenescent mice. Male and female C57BL/6J mice, 85 weeks old, were randomized between two complete purified diets differing only in oil source (cold-pressed safflower or flaxseed). After 8 weeks, humoral competence was assessed in six mice per group as the serum haemagglutinin titre to sheep red blood cells (SRBC) and cell-mediated competence was assessed, in an additional six mice per group, as the delayed hypersensitivity response to SRBC. A zero-time control group (88 weeks old) and a young adult positive control group (12 weeks old) were each tested similarly (six per immune response), revealing age-related depression in both antibody and cell-mediated competence at 88 weeks of age. After the 8-week experimental period, the antibody response of the two test groups of geriatric mice remained below the young adult level (P=0·04) and the cell-mediated response of the saffloweroil group also continued to exhibit age-related depression (20% of young adult level, P=0·0002). By contrast, the anti-SRBC delayedhypersensitivity response of the flaxseed group no longer differed from the response of the young adults but exceeded that of the safflower and zero-time control senescent groups (P=0·0002). Depression in primary cell-mediated competence, the most outstanding aspect of immunosenescence, can be addressed by means of a dietary source of 18:3n-3 without longer-chain PUFA.

Eighteen groups of seventy Atlantic salmon (Salmo salar L.) fry (initial mean body weight 0·8 (sd 0·01) g) were fed on semi-purified diets containing graded levels of l-threonine (Thr) in 15 litres aquaria at a temperature of 14·5±1°C. Doses of Thr represented 1, 31, 41, 51, 62, 72, 83 and 93% of its ideal level for optimumprotein deposition. Indispensable amino acids other than Thr were included in the same proportion (on a g/16g N basis) as in the Atlantic salmon fry whole-body carcass. Following 36d of feeding and a 36h fast, fry were killed for whole-body protein and amino acid analysis. Weight gain (r2 0·98), protein accretion (r2 0·97), and Thr accretion (r2 0·97) were linear (P<0·01) functions of Thr intake. Slope of the Thr accretion regression line showed that the efficiency of Thr utilisation above maintenance was 76%. At zero Thr intake, fry lost 5·4mg Thr/kg body weight0·75 per d. The Thr maintenance requirement was 7·2mg/kg body weight0·75 per d and the Thr requirement for growth was 66mg for 1g protein deposition. Increasing doses of Thr resulted in increased (P<0·05) concentrations of histidine and lysine, and decreased concentrations of isoleucine in whole-body protein. The maintenance need for Thr represented 13·4% of the total need for Thr. The data suggest that efficiency of Thr utilisation above maintenance is constant at all levels of Thr intake between 1 and 93% of the level required for optimum protein deposition.

Formulated in terms of protein synthesis (PS) and protein retention (PR), a definition of turnover-related protein retention efficiency (kP) allows the expression kP=[1+(PS/PR)/6]−1, 6 representing the ratio of the energy equivalent of protein to the cost of synthesis. By combining plausible hormonal and cellular control mechanisms of protein (P) growth, it is possible to derive (PS/PR)=[Q{(P/α)−(4/9)Y−1}]−1+1, allowing the calculation of kP by substitution. The symbol α represents the limit value of protein growth, while the term 4/9 derives from the power in the relationship between the concentration of growth factor-related activator in the nucleus and cell volume (cv). Y is the power in the relationship between cv and total tissue protein, and Q represents the proportion of growth factor-activated nuclei in a tissue. The proportion Q can be estimated from simple functions of intake rates or blood growth factor concentrations. Estimates of Y are derived from histological considerations or calculated from experimental observations; Y=1 for multinucleated skeletal muscle fibres and Y=1/3, 1/2, 1/6 on average for mononucleated cell tissues, skin or bone and viscera, respectively. To apply kP to the whole body, an average value of Y=1/2 can be taken. Experimental observations on tissue protein synthesis and breakdown rates yield direct estimates of kP in satisfactory agreement with comparable theoretical predictions.

It is believed that rare earth elements are not absorbed, and thus they are generally used in some mineral absorption studies as a faecal marker. The aim of the present study was to determine the effect of inulin intake and age on dysprosium (Dy) absorption in rats. Eighty male Wistar rats of four different ages (2, 5, 10 and 20 months) were randomised into either a control group or a group receiving 3·75% inulin in their diet for 4d and then 7·5% inulin until the end of the study. The animals were fed fresh food and water ad libitum for 30d. The intestinal absorption of Dy was determined from a 4d (day 21 to day 25) balance study. Mean faecal Dy recovery (%) in the eight groups (3 months control, 3 months inulin, 6 months control, 6 months inulin, 11 months control, 11 months inulin, 21 months control, 21 months inulin) was 94·0 (sd 8·6), 64·8 (sd 10·1), 95·8 (sd 9·4), 81·5 (sd 12·1), 98·4 (sd 9·8), 87·8 (sd 9·5), 97·8 (sd 6·2) and 84·9 (sd 10·9), respectively. Our results showed clearly that dietary inulin intake decreased faecal Dy recovery in all four rat groups, and faecal Dy recovery was significantly higher in the old rats (10 and 20 months) than in the young and adult rats. These results show that the faecal recovery (or intestinal absorption) of Dy may vary greatly with nutritional or physiological states such as inulin intake or age. The use of rare earth elements as a faecal marker should be thus validated under each nutritional or physiological state before being employed in mineral absorption studies.

Phylloquinone is converted into menaquinone-4 and accumulates in extrahepatic tissues. Neither the route nor the function of the conversion is known. One possible metabolic route might be the release of menadione from phylloquinone by catabolic activity. In the present study we explored the presence of menadione in urine and the effect of vitamin K intake on its excretion. Menadione in urine was analysed by HPLC assay with fluorescence detection. Urine from healthy male volunteers was collected before and after administration of a single dose of K vitamins. Basal menadione excretion in non-supplemented subjects (n 6) was 5·4 (sd 3·2) μg/d. Urinary menadione excretion increased greatly after oral intake of the K vitamins, phylloquinone and menaquinone-4 and -7. This effect was apparent within 1–2h and peaked at about 3h after intake. Amounts of menadione excreted in 24h after vitamin K intake ranged, on a molar basis, from 1 to 5% of the administered dose, indicating that about 5–25% of the ingested K vitamins had been catabolized to menadione. Menadione excretion was not enhanced by phylloquinone administered subcutaneously or by 2′,3′-dihydrophylloquinone administered orally. In archived samples from a depletion/repletion study (Booth et al. (2001)Am J Clin Nutr74, 783–790), urinary menadione excretion mirrored dietary phylloquinone intake. The present study shows that menadione is a catabolic product of K vitamins formed after oral intake. The rapid appearance in urine after oral but not subcutaneous administration suggests that catabolism occurs during intestinal absorption. The observations make it likely that part of the menaquinone-4 in tissues results from uptake and prenylation of circulating menadione

Compared with saturated fat, n-3 long-chain PUFA-rich fish oil improves insulin sensitivity in rats. We studied whether n-3 long-chain PUFA could prevent insulin resistance induced by dexamethasone (a glucocorticoid) in healthy human volunteers. A group of eight subjects was studied twice after a 2d dexamethasone treatment, before and after a 3-week supplementation with fish oil (providing daily doses of 1·1g 20:5n-3 and 0·7g 22:6n-3). The subjects were studied during the basal state and over the 6h following an oral glucose load (1g/kg). Plasma glucose fluxes were traced with [6,6-2H2]glucose and [13C]glucose (naturally 13C-enriched corn glucose). Substrate oxidation was obtained from indirect calorimetry. Following fish oil supplementation, plasma glucose fluxes and substrate oxidation were maintained despite a 17% reduction (P<0·05) in the area under the curve of plasma insulin response, suggesting an insulin-sensitizing effect.

The aim of the study was to characterize the time course of the development of high-fat diet-induced hepatic steatosis and its relation to body fat accretion and changes in plasma lipid profile. Female Sprague–Dawley rats were high-fat fed (HF; 42%, kJ) for 1, 2, 4, 6, 12 and 16 weeks and compared to standard fed rats (SD). Data obtained from HF rats were further analysed by classifying the animals into obesity-prone and obesity-resistant. In HF rats, liver lipid content increased rapidly by approximately 200% during the first 2 weeks, decreased almost to baseline levels between weeks 2 and 6, and re-increased by 17% between weeks 6 and 16 (P<0·05). Body weight, body fat accretion, plasma leptin, NEFA and glycerol concentrations were higher in HF than in SD rats (P<0·05). These higher values were established in 2 weeks and the differences between the groups did not further enlarge from weeks 2 to 16. Obesity-prone rats depicted higher body weight and body fat accretion than obesity-resistant and SD rats. Surprisingly, however, liver lipid content was the same in obesity-prone as in obesity-resistant rats as they were both higher than in SD rats (weeks 2 and 16; P<0·05). Our data support the hypothesis that the liver acts as a systemic buffer, largely increasing its lipid content in the early stage of high-fat feeding. Our results also suggest that the development of non-alcoholic hepatic steatosis is more linked to dietary fat ingestion than to body weight gain.

Alkylresorcinols (AR) are phenolic lipids found in high amounts in wholegrain wheat and rye. They may be important as bioactive components and/or biomarkers of wholegrain wheat and rye intake. AR absorption was studied with pigs during a habitual diet and after a single meal of AR-rich diet. In study 1, four pigs were fed wholegrain rye (RD) and white wheat bread (WD) for 1 week in a cross-over design. Arterial and venous AR concentrations were determined over 540min after feeding on days 5 and 7. In study 2, four pigs were fed a synthetic, starch-based diet for 5d, and arterial and venous AR concentrations were determined over 960min after feeding a single meal of RD. In study 1, plasma AR concentrations after consuming RD peaked at 120min, remained elevated between 120 and 420min, and were re-elevated on refeeding at 480min. In study 2, AR appeared in the plasma after 60min, peaked at 180–240min, and remained elevated after 960min. The apparent elimination half-life was 3·93h, and the maximum plasma concentration was 666 (sem 35) nmol/l. Arterial and venous plasma AR concentrations were similar, suggesting that AR are absorbed via the lymphatic system. AR concentrations in pig plasma increase rapidly after a meal containing AR, and persist in plasma for at least 5d, indicating that AR may be a useful biomarker of wholegrain wheat and rye intake.

We have recently shown that meal-induced insulin sensitization (MIS) occurs after feeding and decreases progressively to insignificance after 24h of fasting and is caused by action of a hepatic insulin sensitizing substance (HISS). In order to carry out quantitative studies of MIS, some standardized meal intake is required. Our objective was to establish animal models to be tested in both the conscious and anaesthetized state using intragastric injection of liquid meals in order to quantify MIS. Insulin sensitivity was assessed before and 90min after the meal using the rapid insulin sensitivity test (RIST) which is a transient euglycaemic clamp. Rats tested in the conscious state were instrumented under anaesthesia 6–9d prior to testing with catheters in the carotid artery, jugular vein and stomach. Meals, injected into the stomach, consisted of a liquid mixed meal, sucrose, glucose or water. The glucose sequestration in response to insulin increased by 90% and 61% following the liquid mixed meal (10ml/kg) in conscious and anaesthetized rats, respectively. Glucose, sucrose and water did not effectively activate MIS. MIS was completely reversed in the conscious model by atropine and completely prevented from developing in the anaesthetized model that had previously undergone hepatic denervation. Gastric administration of a liquid mixed meal but not glucose or sucrose is capable of activating MIS for purposes of mechanistic studies and quantification of the MIS process. The feeding signal is mediated by the hepatic parasympathetic nerves.

Human pregnancy is associated with increased requirements for dietary energy and this increase may be partly offset by reductions in physical activity during gestation. Studies in well-nourished women have shown that the physical activity level (PAL), obtained as the total energy expenditure (TEE) divided by the BMR, decreases in late pregnancy. However, it is not known if this decrease is really caused by reductions in physical activity or if it is the result of decreases in energy expenditure/BMR (the so-called metabolic equivalent, MET) for many activities in late pregnancy. In the present study activity pattern, TEE and BMR were assessed in twenty-three healthy Swedish women before pregnancy as well as in gestational weeks 14 and 32. Activity pattern was assessed using a questionnaire and heart rate recording. TEE was assessed using the doubly labelled water method and BMR was measured by means of indirect calorimetry. When compared to the pre-pregnant value, there was little change in the PAL in gestational week 14 but it was significantly reduced in gestational week 32. Results obtained by means of the questionnaire and by heart rate recording showed that the activity pattern was largely unaffected by pregnancy. The findings support the following conclusion: in a population of well-nourished women where the activity pattern is maintained during pregnancy, the increase in BMR represents approximately the main part of the pregnancy-induced increase in TEE, at least until gestational week 32.

To exert beneficial effects for the host, for example, improving the intestinal microflora, a probiotic must reach the intestine as a viable strain. These properties must be demonstrated by in vitro as well as in vivo methods. However, only a few well-designed human clinical studies have shown these properties. Lactobacillus johnsonii La1 has been shown to give many beneficial effects for the host, but it is unclear whether a viable strain of L. johnsonii La1 has the effect of improving host intestinal microflora. In the present study, a randomised double-blind placebo-controlled cross-over trial was conducted to elucidate the effect of L. johnsonii La1 on human intestinal microflora. Twenty-two young healthy Japanese women were randomly divided into two groups, and either received fermented milk with L. johnsonii La1 or a fermented milk without L. johnsonii La1 (placebo) daily for 21d. Consumption of the fermented milk: (a) increased total Bifidobacterium and Lactobacillus, and decreased lecithinase-positive Clostridium in the faeces; (b) increased the faecal lactic acid concentrations; (c) decreased the faecal pH; (d) increased the defecation frequency. These changes were stronger than those observed with the placebo. L. johnsonii La1 was identified in all subjects only after the consumption of the fermented milk. These results suggest that L. johnsonii La1 can contribute to improve intestinal microflora with probiotic properties.

The objective of the present work was to assess the relationship between serum Se concentrations and environmental determinants (i.e. lifestyle, social activity, geographic region, urban status, education, familial status, physical activity, BMI, tobacco, and food and alcohol consumption). Baseline results from 13017 subjects (7876 women aged 35–60 and 5141 men aged 45–60) who participated in the SU.VI.M.AX (Supplémentation en Vitamines et Minéraux Antioxydants) study were analysed. Fewer than 2% of the volunteers had a serum Se status under 0·75μmol/l, which has been quoted as the cut-off of biological Se sub-deficiency. Women had significantly lower serum Se concentrations than men (1·09 (sd 0·19) μmol/l (n 7423) and 1·14 (sd 0·20) μmol/l (n 4915), P<0·0001, respectively). Significant differences in serum Se concentrations were observed between geographic areas. In both sexes, the serum Se concentration increased with alcohol, meat and fish consumption, and decreased with smoking. In premenopausal women, the serum Se concentration was higher in contraceptive-pill users than in non-users. In women only, age was associated with increased serum Se concentrations, and obesity (BMI≥30kg/m2) was associated with decreased serum Se levels. In men, we observed a decrease in serum Se concentrations with increased consumption of vegetables and fruits. In conclusion, though few of the volunteers participating in the SU.VI.M.AX study had Se status in the sub-deficiency range, 83% of women and 75% of men had serum concentrations below the value considered optimal for glutathione peroxidase activity. The largest Se associations in both sexes were found with regions, smoking, alcohol, meat and fish consumption. Further studies are needed to understand the difference in Se status between genders.

The importance of nutritional intervention for children with cystic fibrosis (CF) is well recognised. It would be expected that the increase in knowledge over the past decade would be reflected in improvements in nutritional status for the CF paediatric population. The aim of the present paper was to evaluate the nutritional status of children with CF, cross-sectionally and longitudinally. Body cell mass adjusted for gender and size (BCM/Htp) was measured in sixty-four children with CF to represent nutritional status and expressed as a Z-score. The cross-sectional results showed a mean BCM/Htp Z-score of 0·54 (sd 1·21), with males having a slightly higher Z-score than females but with a larger variation. At the initial measurement, only one female and one male were considered sub-optimally nourished. The longitudinal analysis after 2 years showed that the mean population had a significantly decreased BCM/Htp Z-score; however, when each gender was analysed separately, this decrease was significant only in the males. At the final measurement, only two females and three males were considered sub-optimally nourished. It is evident from our results that children with CF are well nourished, with only a small percentage considered malnourished. It appears that nutritional status decreases with age, with this decline being more evident in males. These results signify that although children with CF are better nourished with current treatment support, intervention needs to continue throughout a CF patient's life to counteract the changes that occur with age.

Malnutrition and its impact on clinical outcome may be underestimated in hospitalised elderly as many screening procedures require measurements of weight and height that cannot often be undertaken in sick elderly patients. The ‘Malnutrition Universal Screening Tool’ (‘MUST’)has been developed to screen all adults, even if weight and/or height cannot be measured, enabling more complete information on malnutrition prevalence and its impact on clinical outcome to be obtained. In the present study, 150 consecutively admitted elderly patients (age 85 (sd 5·5) years) were recruited prospectively, screened with ‘MUST’ and clinical outcome recorded. Although only 56% of patients could be weighed, all (n 150) could be screened with ‘MUST’; 58% were at malnutrition risk and these individuals had greater mortality (in-hospital and post-discharge, P<0·01) and longer hospital stays (P=0·02) than those at low risk. Both ‘MUST’ categorisation and component scores (BMI, weight loss, acute disease) were significantly related to mortality (P<0·03). Those patients with no measured or recalled weight (‘MUST’ subjective criteria used) had a greater risk of malnutrition (P<0·002) than those who could be weighed and, within both groups, clinical outcome was worse in those at risk of malnutrition. The present study suggests that ‘MUST’ predicts clinical outcome in hospitalised elderly, in whom malnutrition is common (58%). In those who cannot be weighed, a higher prevalence of malnutrition and associated poorer clinical outcome supports the importance of routine screening with a tool, like ‘MUST’, that can be used to screen all patients.

Pre-processed foods often contain a high percentage of lipid, present as emulsions stabilised with various surface-active agents. The acidic gastric environment can affect the behaviour of such emulsions, modifying the lipid spatial distribution and, in turn, the rate of gastric emptying and nutrient delivery to the gut. The aim of the present study was to use echo-planar magnetic resonance imaging (EPI) to determine the behaviour of model olive oil emulsions during gastric processing. Six healthy male volunteers were intubated nasogastrically on two separate occasions and fed 500ml 15% (w/w) olive oil-in-water, surfactant-stabilised emulsions designed to have identical droplet size distribution and which were either stable or unstable under gastric acid conditions. EPI was used to assess the oil fraction of the intragastric emulsions, gastric emptying and to visualise the spatial distribution of the oil at 10, 30 and 50min postprandially. The in vivo imaging measurements of the oil volume fraction of the emulsions correlated well (r 0·66, acid-stable; r 0·52, acid-unstable) with that assayed in the gastric aspirates. Compared with the acid-stable emulsion, the acid-unstable emulsion in the gastric lumen rapidly separated into lipid-depleted ‘aqueous’ and lipid layers. Phase separation in the acid-unstable meal allowed the oil-depleted component to empty first and more rapidly than the stable emulsion as determined by the gastric emptying curves. These pilot data suggest that gastric processing and emptying of high-fat foods could be manipulated by careful choice of emulsifier.

The objective of the present study was to determine the comparative hypotensive responses to drinks containing predominantly fat and carbohydrate (CHO) in healthy elderly subjects. Using a randomised, cross-over study, the participants, twelve elderly subjects, six of them female (72·2 (sd 5·7) years), were investigated. On three separate days, blood pressure (BP) and heart rate were measured following ingestion of 300ml drinks containing: (1) CHO (75g glucose and 93g Polyjoule (CHO polymer) providing 2732kJ (653kcal)); (2) 88% fat (cream blended with milk providing 2732kJ (653kcal)); (3) water. Systolic BP decreased following the CHO drink (P<0·001) and the high-fat drink (P<0·001) but not water; there was no difference in the magnitude of the decrease between the CHO drink and the drink containing fat (13·4v. 15·6mmHg). However, the onset of the fall was slower after the fat-containing drink (13·0v. 26·5min (P=0·01); area under the curve for 0–30min for CHO drink −6·5v. fat-containing drink 125·4 mmHg×min (P=0·043)). We conclude that ingestion of a high-fat drink results in a comparable fall in BP to a CHO drink although the onset is relatively slower. These observations may have implications for the management of postprandial hypotension.

Although many data are available concerning anticarcinogenic effects of industrial conjugated linoleic acid (CLA), few studies have reported the antitumour properties of CLA mixtures originating from ruminant products. The aim of the present study was to investigate the in vitro antiproliferative effects of beef CLA mixtures on breast, lung, colon, melanoma and ovarian human cancer cell lines. For this purpose, four fatty acid (FA) extracts prepared from beef lipid and varying in their CLA composition, their corresponding purified CLA-enriched fractions, and mixtures of pure synthetic CLA, the composition of which reproduced that of the four selected beef samples, were tested on cancer cell lines. Cancer cells were exposed for 48h to medium containing 100μm-FA and their proliferation was determined by quantifying cellular DNA content (Hoechst 33342 dye). Compared with cells incubated without FA, the number of cancer cells was reduced from 25 to 67% (P<0·0001) following FA treatment. Antiproliferative effects of CLA mixtures varied in magnitude according to the source of FA, the CLA composition and the cell lines. CLA mixtures naturally present in beef inhibited the proliferation of human cancer cell lines, a high content in cis-trans isomers allowing the most important antiproliferative effect. Beef total FA exhibited a greater growth-inhibitory activity than their corresponding CLA-enriched fractions. These results suggested that either beef FA other than beef CLA could possess antiproliferative properties and/or the existence of complementary effects of non-conjugated FA and CLA, which could favour the antiproliferative properties of beef total FA.

Considering the negative effects of secondary hyperparathyroidism (SHPT) in patients with chronic renal failure (CRF), the objective of the present study was to evaluate body composition changes using conventional and vector bioimpedance analysis in patients before and after parathyroidectomy (PTX). Twelve adult patients, mean age 43·4 (sd 12·7) years, were evaluated prior to and 6 months after PTX. Diets were assessed with 3d dietary records, and mean energy, protein, calcium and phosphorus intake were estimated from these inventories. Weight, height, BMI and bioelectrical impedance were measured; and biochemical markers of nutritional status (albumin and total protein) and bone metabolism (calcium, phosphorus and intact parathyroid hormone) were determined. No significant differences were observed in mean energy, protein and phosphorus after surgery. There was a significant increase in calcium intake after PTX (382·3 (sd 209·6) mg to 656·6 (sd 313·8) mg; P<0·05). Mean weight, BMI, conventional bioelectrical impedance measurements, total body fat, lean body mass and total body water were unaffected by surgery. However, the phase angle and reactance significantly increased after PTX (5·0° (sd 1·4) to 5·6° (sd 1·3); 44·1 (sd 15·6) Ω to 57·1 (sd 14·4) Ω, respectively). The high levels of intact parathyroid hormone before surgery had a negative effect on total body fat (r −0·69, P<0·05). After PTX, the mean albumin significantly increased (3·9 (sd 0·4) g/dl to 4·2 (sd 0·6) g/dl; P<0·05). PTX for SHPT is associated with certain changes in laboratory values, dietary intake and body composition. The latter is best seen with bioimpedance vector analysis.

Cross-sectional studies report an inverse association between BMI and serum carotenoid concentration. The present study examined the prospective association between BMI and the serum concentration of five carotenoids in the Coronary Artery Risk Development in Young Adults (CARDIA) study. Serum carotenoids (α-carotene, β-carotene, β-cryptoxanthin, zeaxanthin/lutein, lycopene), BMI, dietary intake, physical activity and dietary supplement use were measured at years 0 and 7 in 3071 black and white male and female participants, who were either persistent smokers or non-smokers. Among non-smokers, year 0 BMI predicted year 7 serum carotenoid levels: obese subjects (BMI ≥30kg/m2) had an average concentration of the sum of four carotenoids (α-carotene +β-carotene + zeaxanthin/lutein +β-cryptoxanthin)that was 22% lower than the concentration among subjects with a BMI of less than 22kg/m2. In contrast, the sum of carotenoids among smokers was only 6% lower. Relationships between BMI and serum lycopene were weak. The change from year 0 to year 7 in serum carotenoids, except for lycopene, was inversely associated with the change in BMI among non-smokers but not among smokers. Parallel findings were observed for BMI and serum γ-glutamyl transferase level. In summary, the observation that BMI predicted the evolution of serum carotenoids during a 7-year follow-up among young non-smoking adults is consistent with the hypothesis that carotenoids are decreased in protecting against oxidative stress generated by adipose tissue, while smokers maintain a minimal level of serum carotenoids independent of adiposity. The results for lycopene were, however, discordant from those of the other carotenoids.