It has been claimed that lipid emulsions with a restricted linoleic acid content can improve the safety of total parenteral nutrition (TPN). The tolerability of TPN and its effects on the metabolism of fatty acids were assessed in this prospective, double-blind, randomised study comparing an olive/soyabean oil long-chain triacylglycerol (LCT) with a medium-chain triacylglycerol (MCT)/LCT; 50:50 (w) based lipid emulsion in two groups (O and M, respectively; eleven per group) of severely burned patients. After resuscitation (48–72 h), patients received TPN providing 147 kJ/kg per d (35 kcal/kg per d) with fat (1·3 g/kg per d) for 6 d Plasma fatty acids, laboratory parameters including liver function tests, and plasma cytokines were assessed before and after TPN. Adverse events encountered during TPN and the clinical outcomes of patients within the subsequent 6 months were recorded. With both lipid emulsions, the conversion of linoleic acid in its higher derivatives (di-homo-γ-linolenic acid) improved and essential fatty acid deficiency did not appear. Abnormalities of liver function tests occurred more frequently in the M (nine) than in the O (three) group (P=0·04, Suissa–Shuster test). Seven patients (four from group O and three from group M) died as a consequence of severe sepsis 3–37 d after completion of the 6 d TPN period. When compared with the surviving patients, those who died were older (P=0·01) and hyperglycaemic at baseline (P<0·001), and their plasma IL-6 levels continued to increase (P<0·04). Although fatty acid metabolism and TPN tolerability were similar with both lipid emulsions, the preservation of liver function noted with the use of the olive oil-based lipid emulsions deserves confirmation.

The objective of this paper is to provide glycaemic index (GI) and glycaemic load (GL) values for a variety of foods that are commercially available in the UK and to compare these with previously published values. Fasted subjects were given isoglucidic (50 or 25 g carbohydrate) servings of a glucose reference at least two to three times, and test foods once, on separate occasions. For each test food, tests were repeated in at least eight subjects. Capillary blood glucose was measured via finger-prick samples in fasting subjects (0 min) and at 15, 30, 45, 60, 90 and 120 min after the consumption of each test food. The GI of each test food was calculated geometrically by expressing the incremental area under the blood glucose response curve (IAUC) of each test food as a percentage of each subject's average IAUC for the reference food. GL was calculated as the product of the test food's GI and the amount of available carbohydrate in a reference serving size. The majority of GI values of foods tested in the current study compare well with previously published values. More importantly, our data set provides GI values of several foods previously untested and presents values for foods produced commercially in the UK.

Meat increases absorption of non-haem iron in single-meal studies. The aim of the present study was to investigate, over a 5 d period, the potential increasing effect of consumption of pork meat in a whole diet on the fractional absorption of non-haem iron and the total absorption of iron, when compared to a vegetarian diet. A randomised cross-over design with 3×5 d whole-diet periods with diets containing Danish-produced meat, Polish-produced meat or a vegetarian diet was conducted. Nineteen healthy female subjects completed the study. All main meals in the meat diets contained 60 g of pork meat and all diets had high phytic acid content (1250 μmol/d). All main meals were extrinsically labelled with the radioactive isotope 59Fe and absorption of iron was measured in a whole body counter. The non-haem iron absorption from the Danish meat diet was significantly higher compared to the vegetarian diet (P=0·031). The mean fractional absorption of non-haem iron was 7·9 (se1·1), 6·8 (se 1·0) and 5·3 (se 0·6) % for the Danish and Polish meat diets and vegetarian diet, respectively. Total absorption of iron was higher for both meat diets compared to the vegetarian diet (Danish meat diet: P=0·006, Polish meat diet: P=0·003). The absorption ratios of the present study were well in accordance with absorption ratios estimated using algorithms on iron bioavailability. Neither the meat diets nor the vegetarian diets fulfilled the estimated daily requirements of absorbed iron in spite of a meat intake of 180 g/d in the meat diets.

In intervention and observational studies, it is necessary to determine the number of blood samples required to estimate the true value of Fe-status measures. The aim of the present study was to determine the number of days for blood sampling required in order to measure the ‘true value’ of five Fe-status parameters in young Fe-depleted women and to investigate the effect of menstrual cycle on these measures. Twelve women (aged 23–30 years), non-anaemic but with low Fe stores, participated in the study. Venous blood samples were collected under standardised conditions on fifteen non-consecutive days during a 5-week period. All blood samples were analysed for Hb, serum ferritin (SF), serum transferrin receptors (sTfR), red blood cell volume distribution width (RDW) and reticulocytes (RET), and body Fe stores were calculated as the ratio between sTfR and SF. No systematic changes were found in the investigated parameters during the study. When analytical variations were accounted for, the day-to-day variations (CV%) were as follows: Hb 2·9 %, SF 8·2 %, RET 26·0 %, RDW 2·4 % and sTfR 8·1 %. Calculating the ‘true value’ with a 5 % significance level and 80 % power showed that one blood sample was sufficient for Hb, SF, sTfR and RDW, whereas seven blood-sampling days were needed for RET. In this study, no significant differences in Fe status were found across the menstrual cycle. The conclusions from this study are valid for studies conducted under similar strict conditions.

The aim of this report is to describe breastfeeding duration and introduction of foods in Swedish infants born 1997–9, in relation to current recommendations. A secondary aim is to examine breastfeeding duration and introduction of certain allergenic foods in allergy-risk families (for whom allergy-preventive advice has been issued). Out of 21 700 invited infants, screening questionnaires were completed for 16 070 infants after delivery. Parents to 11 081 infants completed a follow-up questionnaire regarding breastfeeding and introduction of foods and 9849 handed in detailed food diaries at 1 year of age. The percentages of infants who were exclusively breast-fed at 3, 6 and ≥9 months of age were 78·4, 10·1 and 3·9, respectively. The corresponding percentages for partial breastfeeding were 87·8, 68·9 and 43·6. Gluten-containing foods were introduced to 66 % of infants between 4 and 6 months, as recommended at the time of the study, and one-quarter had stopped breastfeeding when gluten was introduced. More than 90 % of parents introduced the first sample of solid food during months 4–6, as recommended. Fish and eggs had been introduced during the first year in 43 % and 29 %, respectively, of infants with atopic heredity. Exclusive breastfeeding duration and time of introduction of solid foods, including gluten, seemed to have been in line with Swedish recommendations at the time, although gluten was often introduced late, and not during ongoing breastfeeding as recommended. The adherence to allergy-preventive advice was less than optimal in infants with atopic heredity.

The effects of two kinds of Escherichia coli (E. coli) strain, wild-type E. coli W3110 and E. coli nir-Ptac, which has enhanced NO2 reduction activity, on oral CH4 emission and NO3 toxicity in NO3-treated sheep were assessed in a respiratory hood system in a 4×6 Youden square design. NO3 (1·3g NaNO3/kg0·75 body weight) and/or E. coli strains were delivered into the rumen through a fistula as a single dose 30min after the morning meal. Escherichia coli cells were inoculated for sheep to provide an initial E. coli cell density of optical density at 660nm of 2, which corresponded to 2×1010 cells/ml. The six treatments consisted of saline, E. coli W3110, E. coli nir-Ptac, NO3, NO3 plus E. coli W3110, and NO3 plus E. coli nir-Ptac. CH4 emission from sheep was reduced by the inoculation of E. coli W3110 or E. coli nir-Ptac by 6% and 12%, respectively. NO3 markedly inhibited CH4 emission from sheep. Compared with sheep given NO3 alone, the inoculation of E. coli W3110 to NO3-infused sheep lessened ruminal and plasma toxic NO2 accumulation and blood methaemoglobin production, while keeping ruminal methanogenesis low. Ruminal and plasma toxic NO2 accumulation and blood methaemoglobin production in sheep were unaffected by the inoculation of E. coli nir-Ptac. These results suggest that ruminal methanogenesis may be reduced by the inoculation of E. coli W3110 or E. coli nir-Ptac. The inoculation of E. coli W3110 may abate NO3 toxicity when NO3 is used to inhibit CH4 emission from ruminants.

Women of childbearing age are at risk of Fe deficiency if insufficient dietary Fe is available to replace menstrual and other Fe losses. Haem Fe represents 10–15 % of dietary Fe intake in meat-rich diets but may contribute 40 % of the total absorbed Fe. The aim of the present study was to determine the relative effects of type of diet and menstrual Fe loss on Fe status in women. Ninety healthy premenopausal women were recruited according to their habitual diet: red meat, poultry/fish or lacto-ovo-vegetarian. Intake of Fe was determined by analysing 7 d duplicate diets, and menstrual Fe loss was measured using the alkaline haematin method. A substantial proportion of women (60 % red meat, 40 % lacto-ovo-vegetarian, 20 % poultry/fish) had low Fe stores (serum ferritin <10 μg/l), but the median serum ferritin concentration was significantly lower in the red meat group (6·8 μg/l (interquartile range 3·3, 16·25)) than in the poultry/fish group (17·5 μg/l (interquartile range 11·3, 22·4) (P<0·01). The mean and standard deviation of dietary Fe intake were significantly different between the groups (P=0·025); the red meat group had a significantly lower intake (10·9 (sd 4·3) mg/d) than the lacto-ovo-vegetarians (14·5 (sd 5·5) mg/d), whereas that of the poultry/fish group (12·8 (sd 5·1) mg/d) was not significantly different from the other groups. There was no relationship between total Fe intake and Fe status, but menstrual Fe loss (P=0·001) and dietary group (P=0·040) were significant predictors of Fe status: poultry/fish diets were associated with higher Fe stores than lacto-ovo-vegetarian diets. Identifying individuals with high menstrual losses should be a key component of strategies to prevent Fe deficiency.

Western diets containing suboptimal Cu concentrations could be widespread. A link between marginal Cu deficiency and CVD has been suggested. The objective of the present study was to investigate the effect of Cu supplementation on both Cu status and CVD risk factors in healthy young women. Sixteen women with a mean age of 24 (sd 2) years participated in a randomised crossover study of three 4-week periods with 3-week washouts between periods. During each intervention period, subjects received 0, 3 or 6 mg elemental Cu/d as CuSO4 in addition to their habitual diet. Blood samples were taken to assess the effect of supplementation on putative markers of Cu status. The content of plasma lipids, lipoprotein (a), apo and certain haemostatic factors, as putative indices of CVD, was also analysed. Daily supplementation with 3 mg Cu significantly increased (P<0·05) serum Cu concentration and the activity of erythrocyte superoxide dismutase, although there was no further significant increase after an intake of 6 mg Cu/d. The concentration of the fibrinolytic factor plasminogen activator inhibitor type 1 was significantly reduced (P<0·05) by about 30% after supplementation with 6 mg Cu/d. No other marker of Cu status or CVD risk factor was affected by Cu supplementation. The results indicate that supplementation with 3 or 6 mg Cu/d may improve Cu status in these healthy young women. Increased Cu intake could reduce the risk of CVD and atherosclerosis in man by promoting improved fibrinolytic capacity.

The relationship between malnutrition and social support was first suggested in the mid-1990s. Despite its plausibility, no empirical studies aimed at obtaining evidence of this association could be located. The goal of the present study was to investigate such evidence. A case–control study was carried out including 101 malnourished children (weight-for-age National Center for Health Statistics/WHO 5th percentile) aged 12–23 months, who were compared with 200 well-nourished children with regard to exposure to a series of factors related to their social support system. Univariate and multiple logistic regressions were carried out, odds ratios being adjusted for per capita family income, mother's schooling, and number of children. The presence of an interaction between income and social support variables was also tested. Absence of a partner living with the mother increased risk of malnutrition (odds ratio 2·4 (95 % CI 1·19, 4·89)), even after adjustment for per capita family income, mother's schooling, and number of children. The lack of economic support during adverse situations accounted for a very high risk of malnutrition (odds ratio 10·1 (95 % CI 3·48, 29·13)) among low-income children, but had no effect on children of higher-income families. Results indicate that receiving economic support is an efficient risk modulator for malnutrition among low-income children. In addition, it was shown that the absence of a partner living with the mother is an important risk factor for malnutrition, with an effect independent from per capita family income, mother's schooling, and number of children.

Exercise prior to eating has repeatedly been shown to reduce postprandial lipaemia. The aim of the present study was to investigate whether this effect was manifest in the presence of two factors that independently mitigate postprandial lipaemia: eumenorrhoea and moderate fat intake. Eight healthy eumenorrhoeic rowers consumed a meal of moderate fat content (35% total energy, 0·66g/kg body mass) 14h after having either rowed at 55% of maximal aerobic power (81% of maximal heart rate) for 80min or rested. Both trials were performed during the luteal phase. Blood samples were drawn before the meal and for 8h postprandially for the measurement of individual fatty acids in the triacylglycerol and NEFA fractions, as well as of glucose, insulin and oestradiol. Plasma oestradiol concentrations were not significantly different in the two trials. The postprandial lipaemic response, expressed as either plasma triacylglycerol concentration or area under the triacylglycerol–time curve, was 35% lower (P<0·05) after exercise. The postprandial glycaemic and insulinaemic responses were also lower, indicating increased insulin sensitivity, whereas the NEFA response was higher, suggesting a lower entrapment of dietary fatty acids in adipose tissue after exercise. Finally, exercise increased the proportion of unsaturated:saturated NEFA during the postprandial period. In conclusion, aerobic exercise performed 14h before a moderate-fat meal reduced postprandial lipaemia in women in the luteal phase. This effect shows the potential of exercise to mitigate even moderate lipaemic responses in eumenorrhoeic women.

We describe a clinical trial to study the efficacy of a casein hydrolysate, prepared using an Aspergillus oryzae protease, containing the major angiotensin-I-converting enzyme inhibitory peptides Val-Pro-Pro (VPP) and Ile-Pro-Pro (IPP) in a single-blind, placebo-controlled study. A total of 131 volunteers with high-normal blood pressure and mild hypertension were randomly divided into four groups (n 32 or 33 in each group). Each volunteer was given two tablets containing four different dosages of VPP and IPP (VPP+IPP: 0, 1·8, 2·5 and 3·6 mg), daily for 6 weeks. A significant decrease in systolic blood pressure was observed at 6 weeks in the active group receiving 1·8 mg (P<0·01) VPP and IPP; in the active groups receiving either 2·5 mg or 3·6 mg, systolic blood pressure was decreased at both 3 weeks (P<0·05 and P<0·05) and 6 weeks (P<0·001 and P<0·0001) compared with systolic blood pressure measured before treatment. Changes in the systolic blood pressure after 6 weeks of treatment in the four groups were −1·7, −6·3, −6·7 and −10·1 mmHg, and these effects were dose dependent. In addition, a significant difference in systolic blood pressure between the placebo group and the VPP and IPP group receiving 3·6 mg was observed (P<0·001) by two-way ANOVA. The antihypertensive effect was greater in mildly hypertensive subjects (n 20 or 21 in each group) than in any of the other subjects. No significant change of diastolic blood pressure was observed for all the test groups, and no differences in diastolic blood pressure in the test sample groups compared with the placebo group were observed during the test period.

Insulin use is common in type 2 diabetes and is frequently accompanied by weight gain, the composition of which is poorly understood. The present study evaluates insulin-induced body composition changes. Body weight and composition of thirty-two type 2 diabetic patients undergoing their first 12 months of insulin therapy were compared with those observed in thirty-two type 2 diabetic patients previously treated on insulin (minimum 1 year). Body composition was determined by simultaneous body water spaces (bioelectrical impedance analysis) and body density measurements. After 6 months, glycosylated Hb (HbA1c) significantly improved in the newly treated group (P<0·0001), but remained stable in those treated previously. HbA1c did not differ between 6 and 12 months in the two groups. Body weight significantly (P=0·04) changed over 12 months in those newly treated only (+2·8 kg), essentially comprising fat-free mass (P=0·044). Fat mass remained unchanged (P=0·85) as did total body water, while extracellular: total body water ratio tended to increase in those newly treated (P=0·059). Weight changes correlated with HbA1c changes (R2 0·134, P=0·002) in the initial 6 months only. Insulin therapy leads to weight gain (2·8 kg), predominantly fat-free mass, over 12 months. After 6 months, newly treated patients continued gaining weight despite an unchanged HbA1c, suggesting the potential anabolic role of insulin in subsequent gains. Therefore, in the initial 6 months, weight gain can be attributed to a ‘glucose control-related effect’ and further gain appears to be due to a ‘non-glucose control-related’ effect of insulin treatment.

The present study investigated whether consuming dairy products naturally enriched in cis-9, trans-11 (c9, t11) conjugated linoleic acid (CLA) by modification of cattle feed increases the concentration of this isomer in plasma and cellular lipids in healthy men. The study had a double-blind cross-over design. Subjects aged 34–60 years consumed dairy products available from food retailers for 1 week and then either control (0·17 g c9, t11 CLA/d; 0·31 g trans-vaccenic acid (tVA)/d) or CLA-enriched (1·43 g c9, t11 CLA/d; 4·71 g tVA/d) dairy products for 6 weeks. After 7 weeks washout, this was repeated with the alternate products. c9, t11 CLA concentration in plasma lipids was lower after consuming the control products, which may reflect the two-fold greater c9, t11 CLA content of the commercial products. Consuming the CLA-enriched dairy products increased the c9, t11 CLA concentration in plasma phosphatidylcholine (PC) (38%; P=0·035), triacylglycerol (TAG) (22%; P<0·0001) and cholesteryl esters (205%; P<0·0001), and in peripheral blood mononuclear cells (PBMC) (238%; P<0·0001), while tVA concentration was greater in plasma PC (65%; P=0·035), TAG (98%; P=0·001) and PBMC (84%; P=0·004). Overall, the present study shows that consumption of naturally enriched dairy products in amounts similar to habitual intakes of these foods increased the c9, t11 CLA content of plasma and cellular lipids.

Current research into dietary factors contributing to the development of allergic diseases is directed towards new active approaches instead of passive elimination diets. The present study aimed to investigate the explanatory role of the diet in a probiotic intervention study on the appearance of atopic eczema (AE) in childhood and the safety of perinatal supplementation with probiotics (Lactobacillus rhamnosus strain GG; ATCC 53 103). A prospective follow-up study from birth to 48 months of children (n 159) with a family history of allergic disease was carried out. Outcome measures included growth, dietary intake assessed with 4 d food diaries and their association with AE by logistic regression models. Increased intakes of retinol, Ca and Zn, with perinatal administration of probiotics, reduced the risk of AE, whilst an increase in intake of ascorbic acid increased the likelihood of AE. Perinatal administration of probiotics was safe, as it did not influence the height (mean difference 0·04 (95 % CI −0·33, 0·40) sd scores, P=0·852) or the weight-for-height (mean difference −3·35 (95 % CI −7·07, 0·37)%, P=0·077) of the children at 48 months with and without perinatal administration of probiotics. Up to 48 months, AE did not affect height (mean difference −0·05 (95 % CI −0·42, 0·33) sd scores, P=0·815), but mean weight-for-height in children with AE was −5·1 % (95 % CI −8·9, −1·2 %) lower compared with children without (P=0·010). The joint effects of nutrients and probiotics need to be considered in active prevention and management schemes for allergic diseases.

The contents of short-chain fatty acids were investigated in rats fed lactitol, lactulose and four fructo-oligosaccharides of different degree of polymerisation and solubility. Fructo-oligosaccharides with a low degree of polymerisation (2–8) generated the highest levels of butyric acid all along the hindgut, whereas fructo-oligosaccharides with a high degree of polymerisation (10–60) generated the highest levels of propionic acid. These specific differences were also generally reflected in the caecal pools and molar proportions of short-chain fatty acids. The lower solubility of the fructo-oligosaccharides was related to a lower degree of caecal fermentation. Lactulose and lactitol yielded high proportions of acetic acid and low proportions of butyric acid. It is concluded that both the degree of polymerisation and the solubility may affect short-chain fatty acid formation, whereas the fructose content per se seem to be of less importance. This may be of interest when designing foods with specific health effects.

The biokinetics of newly absorbed vitamin E in blood components was investigated in normolipidaemic males. Subjects (n 12) ingested encapsulated 150 mg 2H-labelled RRR-α-tocopheryl acetate with a standard meal. Blood was collected at 3, 6, 9, 12, 24 and 48 h post-ingestion. 2H-Labelled and pre-existing unlabelled α-tocopherol (α-T) was determined in plasma, lipoproteins, erythrocytes, platelets and lymphocytes by LC–MS. In all blood components, labelled α-T concentration significantly increased while unlabelled decreased following ingestion (P<0·0001). Significant differences in labelled α-T biokinetic parameters were found between lipoproteins. Time of maximum concentration was significantly lower in chylomicrons, while VLDL had a significantly greater maximum α-T concentration and area under the curve (AUC) (P<0·05). The largest variability occurred in chylomicron α-T transport. Erythrocyte labelled α-T concentrations increased gradually up to 24 h while α-T enrichment of platelets and lymphocytes was slower, plateauing at 48 h. Platelet enrichment with labelled α-T was biphasic, the initial peak coinciding with the labelled α-T peak in chylomicrons. Erythrocyte and HDL AUC were significantly correlated (P<0·005), as was platelet and HDL AUC (P<0·05). There was a lower turnover of pre-existing α-T in platelets and lymphocytes (maximum 25 % labelled α-T) compared to plasma and erythrocytes (maximum 45 % labelled α-T). These data indicate that different processes exist in the uptake and turnover of α-T by blood components and that chylomicron α-T transport is a major determinant of inter-individual variation in vitamin E response. This is important for the understanding of α-T transport and uptake into tissues.

The prenatal diet can program an individual's cardiovascular system towards later higher resting blood pressure and kidney dysfunction, but the extent to which these programmed responses are directly determined by the timing of maternal nutritional manipulation is unknown. In the present study we examined whether maternal nutrient restriction targeted over the period of maximal placental growth, i.e. days 28–80 of gestation, resulted in altered blood pressure or kidney development in the juvenile offspring. This was undertaken in 6-month-old sheep born to mothers fed control (100–150 % of the recommended metabolisable energy (ME) intake for that stage of gestation) or nutrient-restricted (NR; 50 % ME; n 6) diets between days 28 and 80 of gestation. Controls were additionally grouped according to normal (>3, n 7) or low body condition score (LBCS; <2, n 6), thereby enabling us to examine the effect of maternal body composition on later cardiovascular function. From day 80 to term (approximately 147 d) all sheep were fed to 100 % ME. Offspring were weaned at 12 weeks and pasture-reared until 6 months of age when cardiovascular function was determined. Both LBCS and NR sheep tended to have lower resting systolic (control, 85 (SE 2); LBCS, 77 (SE 3); NR, 77 (SE 3) mmHg) and diastolic blood pressure relative to controls. Total nephron count was markedly lower in both LBCS and NR relative to controls (LBCS, 59 (SE 6); NR, 56 (SE 12) %). Our data suggest that maternal body composition around conception is as important as the level of nutrient intake during early pregnancy in programming later cardiovascular health.

The present study compared the nutritional status of schoolchildren from recently settled, ethnic minority tribespeople with those from a Persian village in southern Iran. Height and weight were measured and blood was collected from school children at three time points over 1·5 years. Supplemental Fe was provided to children with low Hb after the first screening. Twenty-one per cent of the children were wasted, 57 % were stunted and 23 % were anaemic. No statistically significant difference in the prevalence of wasting, stunting and anaemia was found between gender or ethnic groups. Children over the age of 12 years had a higher prevalence of wasting than children aged below 12 years. In a sub-sample of forty-one children the average BMI-for-age decreased. Fe supplementation increased Hb levels to normal in most children, but did not increase Fe level in a few children. Dietary deficiency of micronutrients, especially Zn and Fe, probably accounts for the high prevalence of stunting and anaemia in these children. Infection with Helicobacter pylori is another possible explanation for the Fe-deficiency anaemia. Further investigation is in progress to determine the cause(s) of the observed deficiencies.

The role of dietary glycaemic index (GI) and glycaemic load (GL) in disease aetiology is of increasing interest. However, nutritional factors related to dietary GI and GL are not well understood from a population perspective. We aimed to investigate the relation ship between GI and GL and dietary intake at the food and nutrient level. Study subjects were 1071 non-diabetic adults from the Insulin Resistance Atherosclerosis Study, Exam I, 1992–4. Usual dietary intake was assessed with a 114-item modified Block food frequency questionnaire. Published GI values were assigned to food line items. Correlation and regression analyses were conducted. Intake of white bread, beer, meats and fries/fried potatoes was positively associated with average GI, as was fat, starch and alcohol intake (before and after energy adjustment). Intake of fruits and low-fat milk was inversely associated with GI, as were intakes of mono- and disaccharides, and fibre. GL was positively correlated with carbohydrate foods and inversely with non-carbohydrate foods. Gender-specific regression models identified eight food groups explaining 63 % (men) and 55 % (women) total GI variation after adjusting for demographics; 70 % of variation in GL was explained by eleven (men) and nine (women) food groups, respectively. Although the GI of a food is an indicator of the ability of carbohydrates to raise blood glucose, dietary GI, unlike GL, appears to reflect more dimensions of diet than just carbohydrates, such as the combination of foods consumed. This may have implications for the interpretation of dietary GI in epidemiologic studies.

Overall dietary patterns have been associated with health and longevity. We used principal component (PC) and cluster analyses to identify the prevailing dietary patterns of 99 744 participants, aged 60 years or older, living in nine European countries and participating in the European Prospective Investigation into Cancer and Nutrition (EPIC-Elderly cohort) and to examine their socio-demographic and lifestyle correlates. Two PC were identified: PC1 reflects a ‘vegetable-based’ diet with an emphasis on foods of plant origin, rice, pasta and other grain rather than on margarine, potatoes and non-alcoholic beverages. PC2 indicates a ‘sweet- and fat-dominated’ diet with a preference for sweets, added fat and dairy products but not meat, alcohol, bread and eggs. PC1 was associated with a younger age, a higher level of education, physical activity, a higher BMI, a lower waist:hip ratio and never and past smoking. PC2 was associated with older age, less education, never having smoked, a lower BMI and waist:hip ratio and lower levels of physical activity. Elderly individuals in southern Europe scored positively on PC1 and about zero on PC2, whereas the elderly in northern Europe scored negatively on PC1 and variably on PC2. The results of cluster analysis were compatible with the indicated dietary patterns. ‘Vegetable-based’ and a ‘sweet- and fat-dominated’ diets are prevalent among the elderly across Europe, and there is a north–south gradient regarding their dietary choices. Our study contributes to the identification of groups of elderly who are likely to have different prospects for long-term disease occurrence and survival.