We assessed total body water (TBW) and extracellular water (ECW) in thirty-four non-ascitic cirrhotics and twenty healthy controls by 2H2O and Br dilution. In the same subjects, bioelectric impedance (BI) was recorded at multiple frequencies. Body hydration was similar for controls (mean 55·6 (SD 6·7)), less-severe cirrhotics (Child-Pugh classification A; CPA; n 21, mean 56·2 (SD 6·2)) and moderately-severe cirrhotics (Child-Pugh classification B; CPB; n 13, mean 57·2 (SD 5·4)). However, intracellular water standardized per litre TBW was significantly higher in CPB subjects (mean 27·0 (SD 7·5); P<0·01) compared with CPA (mean 21·3 (SD 10·6)) and control subjects (mean 18·0 (SD 9·8)). Published formulas for predicting TBW and ECW from BI at multiple frequencies were applied to the cirrhotics. These formulas gave accurate predictions of TBW and ECW, although standard errors of estimates were higher for CPB subjects (TBW ≤ 2·5 and ECW ≤ 2·11) than those for CPA (TBW ≤ 2·0 and ECW ≤ 1·81) and control (TBW 1·4 and ECW 0·9 1) subjects

It has been suggested that energy expenditure is higher in subjects consuming reduced-fat, high-carbohydrate diets than in subjects consuming full-fat, low-carbohydrate diets. In a 6-month randomized, controlled trial, seventeen women and twenty men (age 20–35 years; BMI 22–28 kg/m2)had free access either to a range of about forty-five reduced-fat products or the full-fat equivalents. At the end of the 6 months, energy intake, sleeping metabolic rate (SMR), average daily metabolic rate (ADMR), and physical activity (AO) were measured. The intervention resulted in a mean difference of the change of the fat content of the diet of 6% of energy (P < 0·01) between the two groups. SMR, ADMR and AO were virtually the same in both groups. The results suggest that the change in fat content of the diet has no effect on physical activity and energy expenditure. However, subjects with a higher activity level consumed more carbohydrate (ADMR/SMR: r = 0·49, P < 0·01; AO: r = 0·57,p=0·001)

For the first time, body composition has been studied in a very large sample of female anorexia nervosa patients (n 200) using two methods: anthropometry (skinfold thickness) and densitometry (underwater weighing). The concurrent validity of both methods appeared to be good (r 0·84). Although the mean percentage of body fat (13·5) in our sample corresponds with most of the previous reports, the range (4·3–24·8) is great. Our study shows that BMI is not a good measure of fatness. Age and duration of illness were not related to body fat. Percentage fat was found to be different (t 2·76, P <0·01) according to the diagnostic subtype of anorexia nervosa: restricters (12·9) v. bingers and/or purgers (14·7).

In view of the importance of long-chain polyunsaturated fatty acids (LCP) for growth and development of fetal and infant neural tissue, the influence of the dietary n-3 and n-6 LCP intake onthe LCP status of forty-three preterm infants (birth weight<1800 g) was studied. Thirty-one formula-fed infants were randomly assigned to receive a conventional formula lacking LCP (n 16), or an 22:6n-3-and 20:4n-6-enriched formula (n 15); twelve infants received their own mother's breast milk. Fatty acid compositions of plasma and erythrocyte (RBC) phospholipids (PL) were determined in umbilical venous blood, in weekly postnatal samples until day 35 of life and, for the formula-fed infants, at 3 months of corrected age. Both in plasma (P < 0·001) and RBC (P < 0.01) PLY, the changes with time until day 35 for 22: 6n-3 and 20:4n-6 in the two groups of formula-fed infants were significantly different, with higher values, comparable with those of human-milk-fed infants, in the LCP-enriched-formula group. At 3 months of corrected age, differences between the two formula-fed groups were even more pronounced. In conclusion, adding 22: 6n-3 and 20:4n-6 to artificial formulas in balanced ratios and in amounts similar to those found in preterm human milk raises both the 22:6n-3 and the 20:4n-6 status of formula-fed preterm infants to values found for human-milk-fed preterm infants. Additional studies are necessary to evaluate the potentially favourable effects of this combined addition on the neurodevelopmental outcome of preterm infants.

Blood examination was conducted for the four Gidra-speaking village groups in Papua New Guinea, who were characterized by high Fe intake and high malaria prevalence with marked inter-village differences. The northern riverine villagers, whose Fe intake was higher than the other three village groups, did not suffer from Fe-deficiency anaemia in their malaria-endemic environment; nor did the inland villagers, with their second highest Fe intake and their malaria-free environment, suffer from Fe-deficiency anaemia. However, several individuals of the southern riverine village suffered from anaemia in a malaria-endemic environment, although their Fe intake was almost the same as the inland villagers’. A considerable proportion of the coastal villagers were anaemic, reflecting the lowest Fe intake and the highest malaria prevalence. An inter-village comparison of the relationships between haemoglobiin levels and transferrin saturation revealed that the southern riverine villagers needed smaller amounts of circulating Fe for erythropoiesis than the northern riverine and inland villagers, reflecting the long-term human-environment conditions such as the density of malaria vectors and the people’s dietary habits. Fe supplementation was not judged effective against hypoferraemia and/or anaemia in such a population. As the incidence of malaria had no significant long-lasting effect on Fe stores or circulating Fe concentration, but did have an effect on anaemia, the hypothesis that malaria causes a transfer of Fe from the blood to parenchymal tissues as a defence against infectious diseases was not supported.

Diarrhoea is a major cause of short-term growth faltering in children of the developing world. If catch-up weight gain is delayed by inadequate dietary intake, or by further bouts of diarrhoea, progressive growth failure occurs. To test the hypothesis that early refeeding is as effective as later feeding after acute diarrhoea with weight loss, we measured the effects of a timed dietary intervention on weight gain after acute diarrhoea in underweight Gambian children. Thirty-four children aged 4–22 months with weight loss following acute diarrhoea were given a high-energy-protein supplement for 14 d beginning either immediately after rehydration or a fortnight later. With a 50% increase in energy intake and a 100% increase in protein intake there was a rapid and highly significant (P < 0.001) gain in weight within a fortnight whether the supplement was given immediately or 2 weeks after presentation. Rates of weight increase were similar whether supplementation was provided early or late, but over the full 28 d (of intervention and non-intervention) children who received late supplementation had greater overall weight gain (P < 0.02) than those supplemented early. Vigorous and early feeding with a high-energy-protein supplement should be central to the management of malnourished children with acute diarrhoea in developing countries, and may be as important as control of diarrhoea in preventing malnutrition and growth failure. This may be achieved in the community using IocaUy available foods, in the face of continuing diarrhoea

Nutrients not absorbed in the small bowel will form substrates for microbial growth in the colon which may have implications for the development of colon cancer. The aim of the present study was to investigate whether fibre-rich oat and barley diets increase the excretion of energy-supplying nutrients from the small bowel compared with a low-fibre wheat diet, and whether a possible. increase could be related to the β-glucan content. Nine ileastomy subjects were served four types of bread together with a low-fibre basal diet (12 g dietary fibre/d). The breads were based on either wheat flour (W diet, 7 g dietary fibre/d), oat bran (OB diet, 29 g dietary fibre/d), the same amount of oat bran with addition of β-glucanase (EC 3.2.1.4) (OBE diet, 19 g dietary fibre/d) or a fibre-rich barley fraction (B diet, 35 g dietary fibre/d). An increased ileal excretion of starch was observed with the barley diet but no effect of the oat β-glucan on starch recovery was found. The NSP + Klason lignin in the ileostomy effluents accounted only for 24, 31, 24 and 35% of the gross energy excretion in the W, OB, OBE and B diet periods respectively. A large part of the dry weight and energy (30, 21, 28 and 27%, in the W, OB, OBE and B diets respectively) in the effluents could not be identified as fat, protein, total starch or NSP + Klason lignin. This unidentified part was probably made up of oligosaccbarides, endogenous losses and nutrient complexes. Methods for identifying and analysing these components should be developed and their role as substrates for colonic fermentation and colon cancer development ought to be investigated.

Hepatic NH3 detoxification by ureagenesis requires an input of aspartate-N, originating either from amino acid-N or NH3-N. The relative importance of these two routes may depend on the nutritional state. To test this, four volunteers were given a liquid diet for 2 d and then on day 3 were either fed every 20 min or fasted. Doses of 15NH4Cl were taken orally every 20 min for 6 h (total 1.5 g) and blood was sampled hourly. Urea-N elimination under fasted conditions was only 0.75 of that for the fed state. Considering the increase in body urea pool during feeding, ureagenesis during fasting was probably closer to 0.6 of that during feeding. Since the [14N15N] urea enrichment was not different between the fed and fasted states, the proportion of the 15NH3 dose converted to urea during fasting was also 0.6 of that during the fed condition. No change in [14N15N] urea and [amide-15N] glutamine enrichment suggested that NH3 enrichment was also not affected by nutritional state. Enrichment of [14N15N] urea was approximately 0.05 that of [14N15N] urea which indicates that 15NH3 can also enter the aspartate route, the importance of which is yet unknown. Both [15N15] urea and [amino-15N]glutamine enrichment in the fasted state were approximately 1.7 times that in the fed state, indicating increased labelling of precursors and/or increased NH3 flux through the aspartate route. Glutamate, valine, leucine and isoleucine showed comparable increases in enrichment during fasting. Arginine enrichment was unaltered by nutritional state, but was lower than [14N15N] urea, indicating incomplete equilibration with the arginine pool in periportal hepatocytes. The present study indicates that hepatic NH3 detoxification may use the aspartate route, gaining importance in the fasted state. The majority of urea was supplied with only one N atom from NH3, thus provision of the other may have consequences for alternative substrates, in particular amino acids.

The aim of the present study was to evaluate the acute effect of ethanol on insulin sensitivity, and glucose, insulin, free fatty acid (FFA), and triacylglycerol responses in ten patients with non-insulin-dependent (type 2) diabetes. In the test study an oral dose of 0·66 g ethanol/kg followed by continuous intravenous infusion of 0·1 g ethanol/kg per h was given to maintain a constant ethanol level in the blood. In the control study identical volumes of oral water and intravenous saline (9 g NaCl/l) were given. After 90 min insulin sensitivity was determined by the hyperinsulinaemic, euglycaemic clamp technique. Ethanol caused no change in blood glucose or insulin concentrations. The FFA level was suppressed by ethanol while the triacylglycerol level was unaffected. The insulin sensitivity was not affected by ethanol. No major acute effect of ethanol on the glycaemic control in fasting type 2 diabetic patients was found in comparison with what is seen in healthy people. The present study, along with the sparse literature, indicates that the ability of ethanol to induce hypoglycaemia is attenuated or absent in diet-treated type 2 diabetes. Furthermore, we found no change in insulin sensitivity. Consequently, the risk of acute ethanol-induced aberrations in carbohydrate metabolism in diet-treated type 2 diabetes seems to be less than previously expected, when alcohol is not taken as a part of a meal.

Vitamin A status of 260 groups of twenty-five males or twenty-five females, aged 35–64 years, surveyed in twenty-four provinces of the People's Republic of China, was assessed by measuring plasma retinol, retinol-binding protein and β-carotene concentrations. Direct measurements of food intake over a 3 d period and questionnaire data on the frequency of consumption of vegetables, fruit, animal products and other dietary items were also used. Vitamin A status appeared to be low only in specific counties but in general was satisfactory or only marginally deficient. Plasma harotene levels were strikingly low in comparison with Western levels despite generous vegetable consumption suggwg that intake of vitamin A precursors may have been adequate but not abundant enough to maintain high circulating plasma levels of β-carotene. Plasma β-carotene, for both males and females, was significantly correlated with the frequency of consumption of green vegetables. Plasma retinol, for males, was highly correlated with meat, fish, oil and alcohol consumption expresPed both in quantity or frequency of consumption. Higher levels of plasma retinol, together with lower levels of plasma β-carotene in males compared with females, suggest that men consume more animal products or may have higher retinol requirements and therefore a higher rate of conversion of β-carotene to retinol.

During the Austral summer of 1992–3, two men, MS and RF, walked 2300 km across Antarctica in 96 d, unassisted by other men, animals or machines. During the journey they ate freeze-dried rations, towed on sledges, that contained an average of 21·3 MJ/d of which 56·7% was fat, 35·5% carbohydrate and 7·8% protein (98·8 g). Despite this high energy intake both men lost more than 20 kg in body weight due to their extremely high energy expenditures. Studies of protein turnover using [15N]glycine by the single-dose end-product method were made before, during and after the journey, and these demonstrated considerable differences in the metabolic responses of the two men to the combined stresses of exercise, cold and undernutrition. However, both men maintained high and relatively stable levels of protein synthesis during the expedition despite the great exertion and the onset of considerable debilitation. This stability indicates the vital physiological function of protein synthesis.

We investigated whether biogenic amines alone, or a combination of NH3, amines and γ-aminobutyric acid (GABA) influenced grass-silage intake, intake behaviour and rumen liquid content in sheep. Three diets were studied: a grass silage preserved with formic acid (4 litres/tonne) (FAS), FAS with 4·9 g amines/kg DM added (FAS+A), and FAS supplemented with a combination of N-components at the following concentrations: 2·7 g amines, 3·0 g NH3and 5·0 g GABA/kg DM (FAS+C). The diets were offered ad libitum, once daily to six rumen-cannulated Texel wethers in a crossover design. Daily DM intake (DMI; g DM/d) tended to be influenced by diet (P = 0·08). The DMI of FAS+A was similar to that of FAS alone, whereas that of FAS+C tended to be higher. The mean rate of ingestion (g DM/min) over all feeding bouts tended to be the lowest for FAS+A (P = 0·06). No differences were found among the diets concerning intake behaviour during the principal meal. Average intake rate of the small meals tended to be the lowest for FAS+A (P = 0·06). Although rumen NH3 concentration was higher (P<0·05) after the principal meal, rumen pH, osmolality, rumen pool size and liquid content were not significantly altered by adding amines or the mixture of N-components to FAS. We conclude that biogenic amines or N-containing products of protein fermentation in concentrations normally found in poor-quality silages do not reduce the intake of well-preserved formic acid-treated silage. A direct effect on chemostatic regulation of intake was not observed, but a slight negative effect on silage palatability cannot be excluded.

Abstract Body composition was measured by densitometry in a group of eighty-one healthy Chinese males and 124 healthy Chinese females, aged 18–67 years. Biceps, triceps, subscapular and suprailiac skinfolds were measured as well as total body bioelectrical impedance at 50 kHz. Mean heights were 1·71 (SD 0·05) and 1·61 (SD 0·06) m, body weights 66·1 (SD 9·1) and 57·2 (SD 9·4) kg and BMI 22·7 (SD 3·0) and 22·0 (SD 3·3) kg/m2 in males and females respectively. Body fat values from body density were 20·1 (SD 6·2) and 29·7 (SD 7·5)% in males and females respectively. Mean predicted body-fat values from skinfolds, bioelectrical impedance and BMI, using prediction formulas developed in Western populations, did not differ or differed only slightly from body-fat values estimated from body density. The SD of the difference for each method was 5% body fat or 3·3 kg fat-free mass (CV 6%), which is approximately equal to the accuracy level of each predictive method. Compared with densitometry the predictive methods overestimated body fat at the lower levels of body fatness, whereas at the higher levels of body fat the predictive methods tended to underestimate body fatness. It is concluded that prediction formulas for estimating body fat from skinfolds, impedance or BMI developed in Western populations are applicable and valid in the adult Chinese population. However, in very lean subjects the predictive methods overestimate body fat compared with values obtained from body density

The effect of long-term supplementation with CaCO3 on indices of Fe, Zn and Mg status was investigated in a randomized, double-blind intervention study of sixty lactating Gambian women. The supplement contained 1000 mg Ca and was consumed between meals 5 d/week, for 1 year starting 1.5 weeks postpartum. Compliance was 100%. Plasma ferritin concentration, plasma Zn concentration and urinary Mg output were measured before, during and after supplementation at 1.5, 13, 52 and 78 weeks postpartum. No significant differences in mineral status were observed at any time between women in the supplement and placebo group. Analysis of the longitudinal data series showed that plasma ferritin and Mg excretion were characteristic of the indivdiual (P < 0·001). Within individuals, ferritin concentration was higher at 1.5 weeks postpartum than later in lactation (P = 0.002). Plasma Zn concentration was lower at 1.5 weeks postpartum than at other tima (P < 0·001), an effect which disappeared after albumin correction. Low plasma concentrations of ferritin and Zn indicated that the Gambian women were at high risk of Fe and Zn deficiency. Measurements of α1-antichymotrypsin suggested that the results were not confounded by acute-phase responses. The results of the present study indicate that 1000 mg Ca as CaCO3 given between meals does not deleteriously affect plasma ferritin and Zn concentrations or urinary Mg excretion in women who are at risk of Fe and Zn deficiency