Proanthocyanidins were prepared from three bean (Vicia faba L.) varieties by extracting hulls in aqueous acetone. The amounts of freeze-dried extracts recovered were 74, 89 and 97 g/kg hull for the varieties Brunette, Statissa and Minica respectively. Chicks (3 weeks old) were fed on a maize–soya-bean control diet or the same control diet substituted with either 30 g proanthocyanidin extracts/kg or 300 g proanthocyanidin-rich hulls/kg. Chicks were tube-fed diets twice daily for 4 d. Nutrient digestibilities were calculated from amounts present in diets and freeze-dried excreta with the aid of titanium dioxide as a marker. Enzyme activities were measured in digesta removed from the jejunum. Extracts of proanthocyanidins depressed the digestibility of protein by 34%, starch by 3% and had no effect on the digestibility of lipid. Proanthocyanidin-rich hulls depressed the digestibility of protein by 62%, starch by 6% and lipid by 4%. Digestive enzyme activities were depressed to the same extent by extracts and hulls, trypsin (EC 3.4.21.4) by 55 and 62%, α-amylase (EC 3.2.1.1) by 75 and 78% and lipase (EC 3.1.1.3) by 31 and 32% for proanthocyanidin-extract and proanthocyanidin-rich-hull diets respectively. The susceptibility of substrates as well as enzymes to the effects of proanthocyanidins is discussed.

The effect of dietary protein on enzyme activity of pancreatic juice was studied in ten growing, castrated, Large White male pigs. Animals, fitted with permanent cannulas in the pancreatic duct and in the duodenum, were divided into two groups receiving either casein or rapeseed concentrate as a protein source. After a 15 d adaptation period to the experimental diet, the volume of pancreatic secretion was significantly higher, whereas the protein concentration was lower in the casein group compared with the rapeseed group. No statistical difference was observed in the daily protein output between groups. Total secreted activities of carboxypeptidase A (EC 3.4.17.1), and elastase (EC 3.4.21.36) were higher in the casein group during the nocturnal period, whereas total activities of trypsin (EC 3.4.21.4), chymotrypsin (EC 3.4.21.1), carboxypeptidase B (EC 3.4.17.2) and amylase (EC 3.2.1.1) in pancreatic secretions during the post-prandial periods were increased by the ingestion of the rapeseed diet. It is concluded that the pancreatic enzyme secretion is sensitive to the nature of the protein ingested.

The objective of the present study was to characterize nutritionally esterase-1 (ES-1). For this purpose, the effects of replacement of dietary carbohydrates by isoenergetic amounts of either fat or protein on ES-1 activities of plasma and small intestine were studied in male rats. Purified diets differing in the amounts of maize starch plus dextrose, casein and various types of fat were used. Plasma and jejunal ES-1 activities were found to be increased with increasing fat intakes. As to the type of fat, increasing plasma ES-1 activities were induced by coconut fat, olive oil, maize oil and medium-chain triacylglycerols, in this order. Maize oil induced higher jejunal ES-1 activities than coconut fat and olive oil, but had similar effects to medium-chain triacylglycerols. Maize oil was more powerful in increasing plasma ES-1 activity than isoenergetic amounts of casein, but with respect to jejunal ES-1 activity these dietary components were equally effective. It is concluded that the amounts of fat and protein in the diet are important determinants of ES-1 activities in plasma and jejunum.

The effect of microbial β-galactosidase (EC 3.2.1.23) activity on intestinal lactose digestion was estimated directly by following post-prandial venous plasma galactose concentrations. To avoid superimposing effects of free galactose, as with yogurt, fresh or heat-treated suspensions of mechanically disintegrated kefir grains in kefir, containing lactose but no free galactose, were fed to ten Göttingen minipigs. Each meal contained 101·1 (SEM 0·1) mmol lactose in kefir supplemented by either native or heat-treated kefir grains corresponding to a mean β-galactosidase activity of either 72 (SEM 8) U or zero. Feeding kefir with β-galactosidase activity resulted in a 30% enhancement of the mean post-prandial plasma galactose peak concentration from 33 (SEM 7) to 43 (SEM 12) μmol/l (n 10), as well as in 23% greater mean areas under the galactose-response curves (8·1 (SEM 1·5) v. 6·6 (SEM 1·2) mmol/min per 1) if compared with kefir with heat-treated grains. Both differences were significant (P < 0·05; paired Wilcoxon test by ranks). There was no induction of intestinal β-galactosidase (EC 3.2.1.108) activity or intestinal lactose-hydrolysing bacteria by lactose feeding. These results give direct evidence of an enhanced lactose digestion and absorption in native fermented milk products due to the microbial β-galactosidase activity.

To test a possible influence of dietary fibre on intestinal sugar uptake, rates of absorption of 10 mm-d[U-14C] glucose and 10 mm-d-[U-14C] xylose were measured in either jejunum or (distal) caecum, by in vivo lumen perfusion, in immature female fowls preconditioned to a standard diet containing (g/kg) either 0, 100, 200 or 400 added dried grass, 200 powdered cellulose, or 200 grass with a polysaccharidase enzyme supplement. When birds were killed after perfusion, dimensions of (unperfused) parts of their alimentary tracts were determined, and recoveries of 14C radioactivity in some body tissues were compared with measured activities absorbed. On average, absorption of glucose was 1·9 and 1·2 times faster than xylose in jejunum and caecum respectively, although these differences varied with dietary treatment and order of perfusion. Increasing grass in the diet caused significant changes in xylose absorption rate in both jejunum and caecum, but only when it was perfused before glucose. With any one sugar and intestinal segment, mean rates of absorption were correlated positively with corresponding mean rates of fluid loss from perfusate. Although their influence on sugar absorption was not well defined, the dietary fibre treatments had more pronounced effects on gross dimensions of parts of the alimentary tract and, hence, potentially on total rates of absorption. Compared with the basal diet, addition of 100, 200 or 400 g grass/kg or 200 g cellulose/kg caused significant increases in small intestine length while 200 g grass/kg with supplementary enzyme did not, and combined caecal length increased with the 400 g grass/kg and the supplementary enzyme treatments. Absorbed 14C activity was recovered in plasma after jejunal perfusions but not caecal ones, whereas it was recovered in liver and in the flushed perfused segment after both types of perfusion. Since there was overlap in absorption rates between jejunum and caecum, this result suggests that the liver may be able to distinguish and treat differently compounds absorbed in the two regions.

Increased intakes of protein have been shown to reduce kidney calcification (nephrocalcinosis) in female rats. Two questions were addressed in the present study. First, can protein-induced inhibition of nephrocalcinosis be demonstrated when the diets used are balanced for calcium, magnesium and phosphorus in the added protein? Second, can the protein effect be explained by the frequently observed magnesiuria after giving high-protein diets? Nephrocalcinosis was induced in female rats by giving purified diets containing 151 g casein/kg and either an increased concentration of P (6 v. 2 g/kg) or a decreased concentration of Mg (0·1 v. 0·4 g/kg). To these diets 151 g ovalbumin/kg was added at the expense of glucose, and the diets were balanced for Ca, Mg and P in ovalbumin. The diets were given for 29 d. In rats fed on the diet containing 151 g protein/kg, an increased intake of P or a decreased intake of Mg caused nephrocalcinosis as measured chemically by analysis of kidney Ca as well as histologically by scoring kidney sections stained according to Von Kossa's method. The addition of ovalbumin to the diet prevented the induction of nephrocalcinosis. High P intake and low Mg intake with the low-protein diets induced enhanced loss of albumin in urine, suggesting that nephrocalcinosis caused kidney damage. Increased protein intake with a non-calcinogenic diet also caused increased albumin excretion in urine. Irrespective of the composition of the background diet, increased protein intake caused increased urinary excretion of Mg. When all dietary groups were considered, differences in nephrocalcinosis and urinary Mg output were not proportionally related.

Nephrocalcinosis: Phosphorus: Magnesium: Protein: Rat

Protein degradation rates were assessed by the excretion of Nt-methylhistidine (NtMH) in four strains of mature chickens, two White Leghorns and two broilers (dwarf and normal), fed on diets containing two levels of dietary protein. Over 0.9 of labelled NtMH was recovered within 7 d of injection from three White Leghorn, three dwarf and three normal broiler males. Protein degradation, measured by NtMH output, was related to adult body-weight by the power 0.71 and strain intercepts were significantly different. Strain differences disappeared when the rate of output of NtMH per unit lean was evaluated. The rate of output of NtMH per unit muscle was higher in birds fed on a low-protein diet of 100 g crude protein (nitrogen x 6.25; CP)/kg compared with males fed on 150 g CP/kg. It was concluded that the lower rate of protein degradation in broiler compared with layer strains at young ages is related to increased adult body-weight in agreement with well-established biological principles.

An experimental model is described that was used for assessing in vivo effects in fowls of exogenous enzyme supplementation on the degradation of plant cell walls to metabolizable monosaccharide residues. It was based on tube-feeding U-14C-labelled cell-wall substrates, cellulose, spinach (Spinacia oleracea) or Festuca with and without enzyme treatments, and monitoring recovery of 14C radioactivity in exhaled carbon dioxide and excreta in the following 8 h. Normal digestion of cell-wall polysaccharides by endogenous microbial activity was also studied by pretreating birds with an antibiotic mixture intended to deplete their intestinal microflora. The results of this pretreatment appeared to confirm the existence of microbial degradation of cellulose in (conditioned) fowls. Judging from differences in 14CO2 production, effects of exogenous enzyme additions were greatly enhanced with all substrates by combining them with a wet pretreatment, thereby increasing the time-period available for them to act in aqueous conditions. However, estimations of digestibilities of cellulose, hemicellulose and pectin with dry and wet treatments, based on recovery of 14C in excreta, indicated that it was only cellulose digestion that was improved by the wet pretreatment. This suggests that degradation of cellulose, which appeared to be slowest, was limited by the dry treatments, whereas that of hemicellulose and pectin was not. Respective digestibilities of these three cell-wall components, from all treatments combined, were in the proportions 1:1.5:4·2.

Plastic particles with different densities (0.92, 1.03, 1.22 and 1.44 g/ml) and sizes (1, 10 and 20 mm) were introduced into the rumen of fistulated sheep kept on a roughage diet. The forestomach was emptied 12 and 24 h after the introduction of the particles, and the contents were replaced by the same amount of rumen contents without plastic particles. The proportions of particles which left the reticulo-rumen (RR) during the experimental period were determined by collecting the faeces during the following 5 d. Non-ruminated particles were separated from the dried RR contents and from the faeces. Large particles were ruminated independently of particle size and density within the investigated range. After 12 and 24 h, 59 and 81% respectively of the particles initially introduced were comminuted due to rumination. During the 12 h period about four times as many particles with a density of 1.44 g/ml passed from the RR into the omasum compared with particles with a density of 0.92 or 1.03 g/ml. Three to ten times more 1 mm particles were excreted than originally-large particles (10 and 20 mm). Particles introduced with an original size of 10 or 20 mm were recovered mostly comminuted in the faeces. In a further experiment the rumens of eight sheep were emptied and filled with a buffer solution. Plastic particles (10 g) of each length (1, 5, 10 and 20 mm; all with a density of 1.03 g/ml) were introduced into the ventral rumen. Sedimentation of particles was prevented by gassing the solution in the RR. Of the initially introduced particles, 31.9, 25.4, 12.7 and 1.5% of the 1, 5, 10 and 20 mm long particles respectively left the RR within 4 h. It is concluded that rumination of particles is independent of particle density and size within the tested range. The probability of particles leaving the RR increases with the higher particle density and with the smaller size. If particle sedimentation is prevented in the RR even 10 mm long particles can leave the RR in considerable amounts.

Particle size: Particle density: Passage: Rumination: Sheep

Four experiments were conducted to compare different methods for estimating the availability of lysine in protein feeds. The same feeds were used in all experiments. In the first experiment the apparent digestibility of lysine to the ileum of growing pigs was found to be 0.92, 0.69, 0.73, 0.85, 0.84 and 0.97 for soya-bean meal, cottonseed meal 1, cottonseed meal 2, meat meal, sunflower meal and skim milk respectively. In the second experiment the utilization of lysine (relative to free lysine) for weight gain, as measured in weaner pigs, was found to be 0.68, 0.73, 0.81, 0.86 and 1.00 for cottonseed meal 1, cottonseed meal 2, meat meal, sunflower meal and skim milk respectively. In Expt 3 diets were formulated to contain soya-bean meal, sunflower meal or cottonseed meal 2 with the same available lysine content from the measurements made in Expts 1 and 2. The feed conversion ratios of weaner pigs given these diets were similar for the three sources of protein when the values were adjusted for the differences in digestible dry matter intake. In the fourth experiment isoenergetic and isonitrogenous diets were formulated with increasing amounts of lysine from synthetic lysine or cottonseed meal 2. The diets were given to weaner pigs and resulted in a curvilinear response to lysine particularly for the cottonseed meal, so that the availability of lysine in cottonseed meal was similar to that of synthetic lysine at low lysine intakes and much lower than synthetic lysine at higher lysine intakes. Possible reasons for the curvilinear responses are discussed.

Eleven fasted adult men consumed a chicken meat sandwich made with white or wholemeal bread, extrinsically labelled with 2 mg 67Zn, on two different occasions. Immediately after eating the sandwich they were given an intravenous injection of 1·5 mg 70Zn. True Zn absorption (which was approximately 7% higher than apparent absorption) was determined by the faecal balance technique by making an allowance for endogenous excretion from measurements of faecal excretion of 70Zn. There was no significant difference in mean true Zn absorption from the white or wholemeal bread sandwich, 33·6 and 25·4% respectively. It was concluded that the substitution of wholemeal for white bread does not reduce Zn absorption from meat-based sandwiches.

A major benefit of supplementing non-ruminant feedstuffs with exogenous enzymes is presumed to be the degradation of plant cell-wall polysaccharides to metabolizable monosaccharide residues. In the present study, metabolic fates of (U-14C-labelled, 10 mM) glucose, galactose, mannose, xylose and arabinose were compared in the fowl, by measuring recoveries of 14C radioactivity in exhaled carbon dioxide, excreta and body tissues after administration either by wing vein (iv) or into the crop (ic). A further comparison was made with a tube-fed, enzyme-treated, U-14C-labelled cell-wall substrate, Festuca arundinacea and a final experiment tested possible competition for absorption between different cell-wall residues. Delays between iv and is treatments in recovery of 14C in CO2, which were assumed to reflect intestinal absorption, indicated that xylose was absorbed more slowly than glucose and galactose, but faster than mannose and arabinose. Total recoveries of 14C in CO2 and excreta over the whole test period indicated that metabolizabilities were highest with glucose, galactose and mannose, and lowest with arabinose. After testing, 14C recovery in caecal contents was highest with is arabinose, and recoveries in body tissues, with all sugar treatments, were in the order liver > breast and leg muscle > abdominal fat > plasma. Results with the Festuca substrate showed similar patterns of recovery in body tissues and confirmed an increase in metabolizability with addition of enzymes. The timing of the 14CO2 response with Festuca and a wet enzyme pretreatment was broadly similar to a ‘predicted Festuca’ response, based on the composition of the substrate and the measured responses with individual (ic) monosaccharides. There was no evidence of any competition for absorption or metabolism among cell-wall residues. It was concluded that glucose release from cellulose is potentially the most important product of cell-wall degradation to contribute to enzyme enhancement of metabolizable energy.

Carboxymethylcellulase (EC 3.2.1.4; CMCase) and xylanase (EC 3.2.1.8) activities were assayed in rumen fluid and from microbes closely associated either with rumen particulate material or with feed particles incubated in nylon bags in the rumen of cattle. The cattle were fitted with a permanent rumen cannula and a simple ‘T’-piece duodenal cannula and were given four diets in a 4 × 4 Latin Square experiment. The basal diet (diet C) consisted of grass silage, barley and rapeseed meal (700, 240 and 60 g/kg total dry matter (DM)) given at the rate of 5·3 kg/d or supplemented with 1·0 kg sucrose/d given twice daily (diet S), twice daily with 0·25 kg sodium bicarbonate/d (diet B) or as a continuous intrarumen infusion (diet I). Giving sucrose supplements decreased CMCase and xylanase activities extracted from microbes associated with rumen particulate material or feed particles incubated in nylon bags as compared with diet C. Supplementation of the sucrose diet with sodium bicarbonate resulted in higher CMCase and xylanase activities than other sucrose diets (S and I). Particle-associated CMCase and xylanase activities were found to be very sensitive in detecting differences in the rumen environment and were related to changes in cell wall digestion. The activities were highly correlated with disappearance of DM and neutral-detergent fibre from nylon bags incubated in the rumen, rumen and total digestion of cell-wall carbohydrates and rumen pool size of cell-wall carbohydrates. It was concluded that the attachment of fibrinolytic enzymes is involved in the depression of fibre digestion. Particle-associated CMCase and xylanase activities were much higher when measured from rumen particulate material than from feed particles incubated in nylon bags.

Randomized, paired studies were carried out on five healthy volunteers equipped with terminal ileostomies to investigate the effect of incorporating 15 g of the viscous polysaccharide guar gum in the normal diet on the volume, weight, composition and physical properties of ileostomy effluent. Subjects ate an identical diet during two 5 d study periods, which were separated by 2 d. Outputs of fat, protein, sodium, potassium, dry weight and water were all increased during guar gum administration, but outputs of carbohydrate, calcium and phosphorus were not significantly altered. Mouth-to-stoma transit was not significantly affected and, surprisingly, the viscosity of the ileostomy effluent was reduced by guar gum. These results show that it is not always possible to predict what will happen to small intestinal function when guar gum is added to the diet from experiments carried out when guar gum is administered alone or with glucose. While our findings show that guar gum will reduce fat absorption, the mechanisms involved are more sophisticated than hitherto envisaged.

The first experiment was a central composite rotatable design with calculated calcium levels of 6.2, 7.0, 9.0, 11.0, and 11.8 g/kg diet and total phosphorus levels of 5.2, 6.0, 8.0, 10.0, and 10.8 g/kg diet (2.8 g phytin-P/kg by analysis). This design involved three replicates for each rotatable point and fifteen replicates of the central point. The second experiment was a 4 × 4 factorial design with calculated Ca levels of 8.0, 10.0, 12.0, and 14.0 g/kg diet and calculated total P levels of 7.0, 9.0, 11.0, and 13.0 g/kg diet (2.5 g phytin-P/kg by analysis). There were four replicates for each treatment. In both 16 d experiments maize–soya-bean diets were used and each replicate consisted of one pen containing 10-d-old broad-breasted, white tom turkeys. The Ca and total P requirements for optimum growth were estimated to be 12.5 and 10.0 g/kg diet respectively. Bone ash was adequate at these levels of Ca and total P, but maximum bone ash was not achieved until much higher levels of Ca and total P were employed. At the required levels of Ca and total P for growth the incidences of Ca- and P-deficiency rickets were very low. There were no treatment effects on feed efficiency. Increasing dietary Ca decreased the incidence of the Ca-deficiency lesion. There was a quadratic response due to dietary total P on both P-deficiency rickets and plasma dialysable P; intermediate levels of dietary P resulted in low incidence of the P-deficiency lesion and high levels of plasma dialysable P. There was a strong negative correlation between the incidence of P-deficiency rickets and plasma dialysable P. Percentage retention was very low at high levels of dietary P and low levels of Ca which corresponded with slightly higher P-deficiency rickets and low plasma dialysable P. No such obvious relationships existed between Ca retention, incidence of Ca-deficiency rickets, and plasma Ca. The incidence of tibial dyschondroplasia was very low in the present study. There were pronounced dietary treatment effects on phytin-P retention; at 14 d percentage phytin-P retention treatment means ranged from 18 to 46 in Expt 1 and from 0 to 40 in Expt 2 with the highest retention of phytin-P at low levels at both Ca and total P.

Calcium: Phosphorus: Phytin-phosphorus retention: Rickets: Tibial dyschondroplasia: Turkey

Effects of quantity and quality of dietary proteins on plasma immunoreactive insulin-like growth factor-1 (IGF-1) concentration, and content of IGF-1 mRNA in rat liver were investigated in rats. Plasma immunoreactive IGF-1 concentration in rats given a casein diet was higher than that in rats given a soya-bean-protein or protein-free diet. The IGF-1 mRNA content in liver was estimated by the Northern blot hybridization technique employing 32P-labelled rat IGF-1 complementary DNA (cDNA). At least four molecular species of IGF-1 mRNA of different molecular weight were found in rat liver. The sizes were 0·8–1·2, 2·0, 3·6–4·0 and 7·4 kb. Most of the mRNA species decreased in the livers of rats given a gluten diet (120 g gluten/kg diet) compared with rats given the casein diet. In particular, mRNA of 7·4 kb decreased markedly. When rats were fed on the protein-free diet, mRNA of all species decreased significantly. The estimated IGF-1 mRNA in the livers of rats fed on the gluten or protein-free diet was almost 0·4 of that of the rats given the casein diet. Feeding the soya-bean-protein diet did not result in a marked effect on the hepatic content of mRNA species of IGF-1. The results showed that liver IGF-1 mRNA content is sensitively regulated by quantity and nutritional quality of dietary proteins.

Two pentosan-rich fractions (water-extractable, WEP, and alkali-extractable, AEP) were isolated from a wheat milling by-product. When both WEP and AEP were added to a commercial-type broiler diet the apparent metabolizable energy (AME), nitrogen retention, feed utilization and growth of broilers were significantly (P < 0.001) depressed. These depressions were closely correlated (P < 0.001) to the level of pentosans in the diets. At the highest level of inclusion of pentosans (equivalent to 40 g arabinoxylans as AEP/kg) the ileal digestibilities of starch, protein and lipid were decreased by 14.6, 18.7 and 25.8% respectively.

Anti-nutritive effects: Non-starch polysaccharides: Broiler diets

Iron speciation in rodent diet and rat gastrointestinal tract lumen during dietary digestion and absorption was investigated with a novel selective extraction technique. Five Fe fractions were identified, namely exchangeable (soluble in 1 M-magnesium chloride), carbonate-bound (soluble in mild acid), oxide-bound (soluble in hydroxylamine–acetic acid), organic-bound (soluble after treatment with peroxide in nitric acid) and residual. Fe from the pelleted diet was mobilized by rat stomach to the exchangeable fraction, then redistributed to the carbonate- and oxide-bound fractions on passage through the proximal small intestine. In vitro incubation of diet with hydrochloric acid failed to mimic the in vivo effect of the stomach. In vitro neutralization of stomach contents with bicarbonate was found to produce a similar effect on Fe speciation to that seen when diet passed the proximal small intestine in vivo. Comparison of 59Fe speciation in extrinsically labelled diet with endogenous Fe speciation showed that extrinsic labelling does not uniformly label all endogenous species. The experiments suggest that selective extraction may provide a useful approach to the study of Fe species present in diets, in vitro digestions and gastrointestinal contents.

A study was carried out to examine whether the responsiveness of small intestinal epithelial cells to dietary carbohydrate varied during the daily 24 h cycle. The effect of sucrose on disaccharidase activities was compared during a period of decreasing disaccharidase activities, i.e. between 22.00 and 10.00 hours, and increasing disaccharidase activities, i.e. between 10.00 and 22.00 hours, in the jejunum of 7-week-old-rats. Rats were fed on a low-starch, high-fat diet (Lst; starch 5 and fat 73% of gross energy), or a high-starch, low-fat diet (Hst; starch 70 and fat 7% of gross energy). Both dietary groups exhibited typical diurnal variations in jejunal sucrase (EC 3.2.1.48), maltase (EC 3.2.1.20) and lactase (EC 3.2.1.23) activities, exhibiting a peak around 22.00 hours and a trough at approximately 10.00 hours. When rats were fed on diet Lst for 7 d and then force-fed on an isoenergetic sucrose diet (S; sucrose 40 and fat 37% of gross energy) for 6 or 12 h they exhibited increased sucrase, maltase and lactase activities compared with rats fed on diet Lst. The absolute increase in disaccharidase activities was similar regardless of the time diet S was given or whether rats were killed at 10.00 hours or at 22.00 hours. Analyses of sucrase and lactase activities along the villus–crypt columns showed that the distribution of cell cohorts that responded to diet S was not influenced by the time of introduction of diet S. These findings suggest that small intestinal epithelial cells possess the ability to respond to dietary carbohydrate throughout the daily 24 h cycle.

Marginally vitamin C-deficient guinea-pigs treated with the diabetogenic agent streptozotocin were compared with those liberally supplied with vitamin C, for functional indices of vitamin C status, particularly in the eye lens. Weanling male Dunkin–Hartley guinea-pigs were fed on diets with 0.1 g vitamin C/kg (marginally deficient), or 5 g/kg (liberally supplied), and some received intraperitoneal streptozotocin (two doses of 150 mg/kg body-weight). About half the streptozotocin-treated animals had high urinary glucose following an oral glucose dose; these animals also grew more slowly than the others. At 4 months after streptozotocin the animals were killed for measurement of tissue vitamin C, glucose and sorbitol. Streptozotocin moderately increased the concentration of glucose in plasma, lens and aqueous humour. Lens sorbitol levels increased only in the group exposed to streptozotocin plus marginal vitamin C. There was a significant (P < 0.02) positive correlation between urinary glucose and lens sorbitol levels overall. Liberal vitamin C intake may thus counteract the effect of streptozotocin diabetes on lens sorbitol, suggesting a new function of vitamin C, possibly related to cataractogenesis and to the biochemical lesions associated with diabetes.