D-Tagatose is an isomer of D-fructose and is produced from lactose in a two-step process. Its properties permit its use as a bulk sweetener, humectant, texturizer, and stabilizer. D-Tagatose has not been evaluated previously by the Committee.

2.1.1 Absorption, distribution, and excretion

Groups of two to four male Sprague-Dawley rats, adapted or unadapted to a diet containing 10% D-tagatose, received a single oral or intravenous dose of D-[U-¹⁴C]tagatose. All rats had free access to tap-water acidified with HCl to pH 2.5 to control microbial growth. Two male, germ-free unadapted rats received a single oral dose of D-[U-¹⁴C]tagatose. All animals were killed 24-72 h after dosing, and radiolabel was determined in expired air, urine, faeces, selected tissues, caecal contents, the intestinal tract with contents, the carcass, and rinsings of the chamber with deionized water. Furthermore, total radiolabel in blood and caecal samples was measured, and these tissues plus urine and faeces were analysed. The results are shown in Table 1.

Unadapted conventional rats killed after 72 h had excreted 49% of the oral dose in expired air, 5.8% in urine, and 8.8% in tissues and carcass; these values represent the absorbed fraction, 64%. In germ-free unadapted rats killed after 24 h, the absorbed fraction of the oral dose was 32%, with 22% in expired air, 3.8% in urine, and 6.2% in tissues and carcass. In adapted conventional rats killed after 72 h, the absorbed fraction represented 83% of the dose, with 68 % in expired air, 5.2% in urine, and 10% in tissues and carcass. The first 6 h after oral dosing are generally regarded as the time during which absorption and metabolism of absorbed nutrients by mammalian enzymes occurs in conventional rats. During this period, ¹⁴C-carbon dioxide accounted for 40% of the oral dose excreted by adapted conventional rats and 14% of that excreted by unadapted germ-free rats, the difference being attributable to microbial fermentation of D-tagatose in the gut of conventional rats. The greater expiration of ¹⁴C-carbon dioxide by adapted than by unadapted conventional rats was ascribed to faster, more complete fermentation of unabsorbed D-tagatose by intestinal microbiota in the adapted rats. The role of adaptation was confirmed by examination of the time course of appearance of ¹⁴C in the expired air of adapted and unadapted rats and by the fact that much less unchanged D-tagatose was recovered from the faeces of adapted rats than from those of unadapted rats.

Table 1. Metabolism of D-[U¹⁴C]tagatose in conventional and germ-free male Sprague-Dawley rats

Only small quantities of intermediate metabolites (lactate, short-chain fatty acids, and pyruvate) were found in blood and urine, suggesting that much of the ¹⁴C-carbon dioxide detected (68% of the oral dose in adapted conventional rats) and possibly ¹⁴C-labelled methane was formed in the gastrointestinal tract and either absorbed into the blood and expired or eliminated as flatus and eructations. Expiration of about 14% of the administered dose as ¹⁴C-carbon dioxide by the germ-free rats within the first 6 h indicates absorption of 20-23%. The finding of ¹⁴C in expired air after both oral and intravenous dosing demonstrates that absorbed D-tagatose is readily metabolized via the glycolic pathway (Saunders et al., 1999a).

Five castrated boars with permanent cannulae in the portal vein, one mesenteric vein, and one mesenteric artery were used to study the absorption of D-tagatose by the small intestine. The boars were fed a diet containing 20% sucrose for 9 days, which was replaced on day 10 by 20% D-tagatose. The animals were fed twice daily, but feed intake was restricted to about 3% of body weight per day, and water was mixed with the feed at a ratio 1:2.5 (feed:water). Each boar was used as its own control. Blood samples were taken simultaneously from the portal vein, mesenteric vein, and mesenteric artery on day 7 (sucrose diet), day 10 (animals unadapted to D-tagatose), and day 17 (animals adapted to D-tagatose) 1 h before morning feeding, every 30 min for 4 h after feeding, and every 60 min thereafter to 12 h after feeding for analysis of D-tagatose, short-chain fatty acids, para-aminohippuric acid, and packed cell volume. Urine was collected during the 12-h blood sampling periods and analysed for D-tagatose.

The concentration of D-tagatose in portal venous blood increased in unadapted and adapted boars shortly after intake, reaching a maximum after 90-150 min. The concentration in arterial blood reached about 75% of that in venous blood, indicating little clearance on first passage through the liver. The concentration of D-tagatose decreased after 150 min and returned to baseline after about 10 h. The average portal absorption of D-tagatose represented 26% and 28% of the ingested dose in the unadapted and adapted animals, respectively. The average amounts of the ingested dose excreted in the urine were 4.7% by unadapted boars and 5.3% by adapted animals, and the average amounts of the D-tagatose absorbed into the portal blood that were excreted in the urine were 18 and 19%, respectively. The net portal absorption of short-chain fatty acids was slightly greater in boars fed D-tagatose than in those fed sucrose in the diet, but no significant difference was seen between unadapted and adapted animals in the absorption of short-chain fatty acids in general or of any individual short-chain fatty acid. Approximately equal amounts of acetic acid were absorbed, but slightly more propionic acid was absorbed by boars given D-tagatose than by those given sucrose. Up to five times more butyric acid was absorbed by boars given D-tagatose than by those given sucrose (Jensen & Laue, 1998).

Two groups of eight castrated boars received a low-fibre diet with 15% sucrose or a similar diet with 100 g/kg sucrose replaced by 100 g/kg D-tagatose, for 17 days. The animals were fed twice daily, but feed intake was restricted to 40 g/kg bw per day. Chromic oxide was used as a marker of digestibility. D-Tagatose was digested and absorbed in the small intestine to only a small extent (26%) and showed some hyperosmotic effect in this part of the gastrointestinal tract. Much of the ingested D-tagatose (> 74%) was completely fermented in the caecum and proximal colon, such that no unchanged D-tagatose was detected in faeces. This contributed to the overall energy balance, with greater production of short-chain fatty acids (in particular, higher concentrations of C3-C5 acids) than in animals given sucrose (Laerke & Jensen, 1999a).

Six pigs from two litters were fed diets containing 20% sucrose, 10% sucrose and 10% D-tagatose, or 20% D-tagatose in a Latin square change-over design. One pig from each litter was fed one of the diets twice a day for 2 weeks and then switched to another diet, for a total of four 2-week intake periods. During the second week, body weight, feed intake, D-tagatose intake, sucrose intake, faecal dry matter, and faecal dry matter output were measured. In addition, the digestibility of dietary nutrients, protein, and fat, the energy balance, the amount and composition of urine, the urinary output of D-tagatose, the amounts of methane and hydrogen expired, and the faecal concentration, composition, and output of short-chain fatty acids and lactate were determined.

Several changes were observed in the group receiving 20% D-tagatose when compared with the other groups:

• Overall faecal output was increased by 46% over that in animals given 20% sucrose and by 41% over that in animals given 10% D-tagatose (30 and 18%, respectively, on dry matter basis).
• The energy yield was 80% with 20% D-tagatose and 85% with 20% sucrose.
• Protein digestibility was decreased.
• The excretion of proteins in urine was decreased, whereas that in faeces was increased.

Analysis of short-chain fatty acids in faeces showed that the molar proportions of propionic acid, butyric acic, and valeric acid had increased and the molar proportion of acetic acid had decreased with increasing D-tagatose intake. About 5% of the ingested D-tagatose was excreted in the urine of pigs given 10% or 20% of this sugar. Only 0.4 g of D-tagatose was found daily in the faeces of pigs given 20% D-tagatose (intake, 360 g/day) and none in those given 10% D-tagatose. Of the ingested D-tagatose, 2% was expired as hydrogen. Slightly less methane was expired by animals given D-tagatose than by those given sucrose, but the finding was independent of D-tagatose intake; however, the amount of methane expired by pigs given D-tagatose was in the same range as those fed a ‘normal’ low-fibre diet. The net energy derived from D-tagatose was estimated to be 1.0-1.6 kcal/g (Jørgensen & Laerke, 1998).

The absorption of D-tagatose by humans was estimated on the basis of data for L-rhamnose, a compound with a slightly lower relative molecular mass and slightly greater lipophilicity than D-tagatose. Since the rate of passive diffusion of a substance through the intestinal mucosa is dependent on its molecular volume and its lipophilicity, the intestinal absorption of D-tagatose was presumed to be somewhat greater than that of L-rhamnose. The absorption of ingested L-rhamnose can be determined quantitatively from its urinary excretion because this sugar is metabolized slowly and thus incompletely. After intravenous administration of L-rhamnose to dogs and humans, about 70% of the dose was excreted in urine. After oral administration to humans, 10-17% of the dose was found in urine. Thus, about 100/70 × (10-17%) = 14-24% of ingested L-rhamnose is absorbed. Since the absorption of D-tagatose is presumed to be somewhat lower, the fractional absorption of D-tagatose by humans is not expected to exceed 20% (from Bär, 1999a).

In contrast to the low absorption of D-tagatose reported above, up to 80% was found to be absorbed in studies of patients with an ileostomy (Normén et al., 1999). Nearly 50% absorption of L-rhamnose was reported in similar studies (Jenkins et al., 1994), and high absorption rates for sorbitol, maltitol, and isomalt were reported in studies in patients with an ileostomy even though these polyols are known to be poorly digested and absorbed. The ileostomy model is therefore of limited value for determining the rate of absorption of malabsorbed monosaccharides and polyols (as cited from Bär, 1999a).

In studies of humans receiving a single oral dose of 20-30 g of D-tagatose, mild intestinal effects (e.g. flatulence) were observed, providing indirect evidence of incomplete absorption (see Observations in humans). In a double-blind, balanced study of cross-over design, eight volunteers received 30 g of D-tagatose or 30 g sucrose per day (at lunch) for 15 days. Metabolic measurements were performed on days 1 and 15.

Total 24-h energy expenditure and hour-by-hour profiles of energy expenditure did not show changes, and no heating effect was seen. However, 24-h hydrogen production showed a 35% increase during administration of D-tagatose, indicating that a substantial portion escapes absorption and is fermented in the large intestine. Adaptation did not influence the hydrogen production (Buemann et al., 1998).

After eight men received a single oral dose of 30 g of D-tagatose or D-fructose in water in a double-blind cross-over study, only 1.5% of the dose of D-tagatose was recovered in urine (Buemann et al., 1999a).

2.1.2 Biotransformation

Incubation of purified bovine liver fructokinase with ATP and with K⁺ and D-tagatose as the substrate resulted in phosphorylation of D-tagatose to D-tagatose-1-phosphate, with a 1.7 times higher (corrected) Michaelis constant (K_m) and about the same V_max as D-fructose (Raushel & Cleland, 1977). The higher K_m for phosphorylation of D-tagatose is probably of no relevance to the intracellular formation of fructose-1-phosphate and D-tagatose-1-phosphate in vivo, because the K_m for cellular uptake of fructose (and probably also of D-tagatose) is substantially higher; i.e. the uptake limits the rate of phosphorylation at physiological concentrations of fructose and D-tagatose < 1 mmol/L (from Bär, 1999a). The presence of D-tagatose-1-phosphate in the liver was demonstrated by ³¹P magnetic resonance spectroscopy in a man who ingested 30 g of D-tagatose. A peak was reached after 30-60 min, and after 150 min D-tagatose-1-phosphate had disappeared (Buemann et al., 1999b). Formation of D-tagatose-1-phosphate probably occurs mainly in the liver and kidney and to a lesser extent in intestinal mucosa and pancreatic islet cells because fructokinase is present only in those tissues (Van den Berghe, 1986; Malaisse et al., 1989; Mayes, 1993).

After eight men received a single oral dose of 30 g of D-tagatose or D-fructose in water in a double-blind cross-over study, D-tagatose caused a transient decrease in inorganic phosphorus in serum, probably due to accumulation of D-tagatose-1-phosphate (Buemann et al., 1999a).

Incubation of D-tagatose with hamster and rat hepatocytes in vitro resulted in the production of glucose, and the rate of gluconeogenesis from 20 mmol/L D-tagatose was about twice that from 20 mmol/L D-galactose and half that from fructose. The distribution of ¹⁴C in the glucose formed with ¹⁴C-fructose and ¹⁴C-D-tagatose was identical, suggesting that the two sugars have the same major pathway of gluconeogenesis. Thus, D-tagatose-1-phosphate is cleaved by aldolase B in the liver to D-glyceraldehyde and dihydroxyacetone-phosphate. Studies with isotopes also showed that addition of fructose to D-tagatose inhibited the conversion of D-tagatose to glucose. Addition of D-tagatose to fructose did not affect gluconeogenesis from fructose (Rognstad, 1975, 1982). The pathway of gluconeogenesis from fructose in the liver is as follows: fructose ^fructokinase > fructose-1-phosphate ^{aldolase B} > dihydroxyacetone-phosphate + D-glyceraldehyde (-> D-glyceraldehyde-3-phosphate) -> fructose-diphosphate -> fructose-6-phosphate -> glucose-6-phosphate -> glucose. However, the pathway of gluconeogenesis from D-tagatose has not been established, and it may be metabolized in mammalian liver via a pathway involving a galactose phosphate; e.g. via an isomerase that catalyses the conversion of D-tagatose-1-phospphate to galactose-1-phosphate (from Bär, 1999a).

Since the rate of gluconeogenesis from D-tagatose was about half that from fructose in isolated rat hepatocytes, cleavage of D-tagatose-1-phosphate occurs at about half the rate of that of D-fructose-1-phosphate (Rognstad, 1975, 1982). Further studies with rat hepatocytes showed a lower affinity of aldolase B for D-tagatose-1-phosphate, as demonstrated by a longer reduction of ATP and inorganic phosphorus concentrations after incubation with D-tagatose than with fructose and less glycolysis with D-tagatose (Martínez et al., 1982, 1987; Vincent et al., 1989). The longer reduction of ATP and inorganic phosphorus concentrations may be a consequence of the slow degradation of D-tagatose-1-phosphate, leading to accumulation. As seen in patients with heriditary fructose intolerance, this may increase the degradation of purine and increase the production of uric acid. Ingestion of 30 g D-tagatose by a human subject decreased the beta-ATP concentration by about 12% and was maximal for 30-60 min, contemporary with the peak of D-tagatose-1-phosphate (Buemann et al., 1999b).

2.1.3 Fermentation in the gastrointestinal tract

Twelve castrated boars were fed diets containing 20% sucrose, 10% sucrose and 10% D-tagatose, or 20% D-tagatose and were killed after at least 2 weeks on the diet. Those fed 20% sucrose and 10% D-tagatose were killed 6 h after the morning feed, while those on the 20% D-tagatose diet were killed 0, 3, 6, or 9 h after the morning feed. The gastrointestinal tracts were then cut into eight segments (stomach, three segments of small intestine, caecum, and three segments of large intestine), and samples were taken and incubated. Addition of D-tagatose to the diet caused increased production of butyric acid and valeric acid in the caecal and colonic contents, the percentage of butyric acid increasing from 18% with sucrose to 31% with 20% D-tagatose and that of valeric acid from 3.4% with sucrose to 9.0% with 20% D-tagatose. The percentage of acetic acid decreased from 51% in animals on the sucrose diet to 30% in those fed 20% D-tagatose. The percentage of propionic acid was unaffected by diet (23% with sucrose, 26% with D-tagatose) (Jensen & Laue, 1998).

In the experiment of Laerke & Jensen described above, faecal samples were taken before the experiment (with standard diet) and on days 1, 8, and 15. When the pigs were killed on day 17, samples were taken from the stomach, the distal third of the small intestine, the caecum and the mid-third of the colon and incubated. The population of degrading bacteria in faeces and the degrading capacity of bacteria in the hindgut were greater in adapted than in undapted boars. Microbial fermentation occurred readily in the caecum and colon but did not occur in the stomach or small intestine. Microbial fermentation of D-tagatose by gut microbiota yielded formate, acetate, propionate, butyrate, valerate, caproate, and some heptanoate and also hydrogen and methane. Acetic acid was the major fermentation product of D-tagatose in unadapted animals. The percentages of butyric and valeric acid produced by microflora in the colon were two to three times higher in adapted than in unadapted animals, and the rate of production of propionic acid by caecal bacteria was more than two times higher in boars given D-tagatose than in the controls. The recovery of energy from D-tagatose as short-chain fatty acids formed by fermentation was estimated to be 40-48% in adapted animals and 33-35% in unadapted animals (Laerke & Jensen, 1999b).

Sixteen healthy people aged 21-36 received 10 g of D-tagatose three times per day for 14 days. Stool samples were collected 1 day before treatment (unadapted) and on day 14 (adapted) and were analysed for their composition of microbiota and for fermentation of D-tagatose and sucrose in vitro. The population density of lactic acid bacteria in the faeces of adapted persons was significantly greater than that of unadapted persons. The population density of coliform bacteria was decreased in the faeces of adapted persons. The major fermentation products of D-tagatose were acetic acid, butyric acid, and caproic acid. The rate of fermentation of D-tagatose was higher in faecal samples from adapted persons, and the resulting short-chain fatty acids had a greater proportion of butyric acid. As estimated from the production and composition of short-chain fatty acids, 57% and 50% of the energy in D-tagatose was recovered from adapted and unadapted faecal samples, respectively (Jensen & Buemann, 1998).

The ability of 176 normal and pathogenic human enteric bacteria and dairy lactic acid bacteria to ferment D-tagatose was studied. A few of the 34 naturally occurring enteric human bacteria were able to ferment D-tagatose, including Clostridium inocuum, Enterococcus faecalis, and two out of three Lactobacillus strains which fermented strongly and Clostridium tertium which fermented weakly. None of the 11 pathogenic human enteric bacteria was able to ferment D-tagatose. Of 22 enteric isolates from normal humans, all four isolates of E. fecalis, one of five of E. faecium, none of five Bifidobacterium strains, and six out of eight Lactobacillus species (four of these strongly) fermented D-tagatose. Most of the 89 dairy lactic acid bacteria fermented D-tagatose extensively, Lactobacillus, Leuconostoc, and Pediococcus strains most strongly, but also strains of Enterococcus, Streptococcus, and Lactococcus. None of the Bifidobacterium strains was able to ferment D-tagatose. Screening of 20 lactic acid bacteria reflected the results for the dairy lactic acid bacteria. E. faecium, Streptococcus thermophilus, and the Lactobacillus species L. casei, L. rhamnosus, L. paracasei, L. fermentum, and a few L. acidophilus showed strong fermentation (Bertelsen et al., 1999).

2.1.4 Effects on absorption of fructose

High concentrations of D-tagatose did not inhibit the initial rate of uptake (unidirectional influx) of D-fructose across the brush border of the rat jejunum in vitro or in vivo. The initial rate of uptake of D-fructose was shown to be a saturable function of concentration and appeared to be based on a carrier mechanism, which was not inhibited by D-tagatose (Sigrist-Nelson & Hopfer, 1974; Crouzoulon, 1978).

2.2.1 Acute toxicity

In a study that did not conform to GLP or current standards, in which the results were given only in summary form with no individual data on clinical observations or macroscopic findings, five male CD-1 mice received D-tagatose as a single oral dose of 10 g/kg bw by gavage as a 75% (w/v) solution in water. The animals were observed clinically for 14 days. Body weights were recorded on days 0 and 14, and the animals were examined macroscopically at termination. No deaths occurred, and the body weights were normal. One mouse showed anogenital staining on days 3 and 4. No abnormalities were seen (Trimmer, 1989).

In a study with the same limitations as that described above, five male and five female Sprague-Dawley rats received D-tagatose as a single oral dose of 10 g/kg bw by gavage as a 75% (w/v) solution in water. The animals were observed clinically for 14 days. Body weights were recorded on days 0 and 14, and the animals were examined macroscopically at termination. One female died immediately after treatment due to a dosing accident. The body weights were normal. One rat showed anogenital staining after 1 h. Enlarged cervical lymph nodes were found in two males and one female, and a discoloured adrenal gland was observed in one female (Trimmer, 1989).

2.2.2 Short-term studies of toxicity

In a study reported only in an abstract, groups of five lean and five obese diabetic rats (SHR/N-cp) were fed a diet containing 24% fructose, 10% glucose, 10% starch, 19% fat, 10% lactalbumin, 10% casein, 5.9% cellulose, 3.1% AIN salt mix, and 1.0% vitamins to which 10% fructose or 10% D-tagatose was added for 2 weeks. The rotation was then repeated. D-Tagatose reduced food efficiency in the lean rats but increased that in the diabetic rats, which also had reduced polydipsia and urinary glucose. These effects were more severe during the second rotation, in which food efficiency was reduced even in the diabetic obese rats and urinary glucose was normal. No pathological changes were observed in any group. The diabetic rats given D-tagatose had an increased calcium concentration in the kidney, probably as a consequence of improved calcium retention due to the absence of polydipsia (Szepesi et al., 1996).

Groups of 30 adult male Sprague-Dawley rats received lab chow to which 0 or 20% D-tagatose was added for 4 weeks. Ten animals per group were killed without a fasting period prior to killing 2, 4, and 6 weeks after the start of treatment. An additional group of 10 animals received a commercial rodent diet (SDS) into which 20% D-tagatose was incorporated at the expense of 20% barley for 4 weeks. The control group received SDS diet into which 20% pregelatinized potato starch was incorporated at the expense of 20% barley. The SDS groups were killed after 4 weeks without a fasting period prior to killing. No statement of compliance with GLP was included.

The groups given 20% D-tagatose in either lab chow or SDS had significantly lower body weights; those of rats on lab chow returned to normal during the 2-week recovery period. No change in body weight was seen in the two control groups. Food intake was reduced during the first week of treatment in groups given 20% D-tagatose in either chow but was increased in those on lab chow during the recovery period. The absolute and relative weights of wet and freeze-dried liver were increased after 2 weeks by 22-26% in the group given lab chow with 20% D-tagatose and were increased by 35-39% after 4 weeks when compared with the control group on the same chow. After the 2-week recovery period, the liver weights were still increased, but the increases were only 9-14%. The liver weights of the group on SDS diet with 20% D-tagatose were increased after 4 weeks to the same extent (by 38-46%) as those of animals given lab chow and D-tagatose. The weights of the liver in both control groups were similar after 4 weeks. The water content of the livers of rats on lab chow with 20% D-tagatose was similar to that of the control group, but the DNA per gram of liver was decreased, indicating cell enlargement. Because the glycogen content was increased (at the expense of protein and total lipid content), the cell enlargement might have been due to glycogen accumulation, but the increased amounts of DNA and protein per whole liver in the group on lab chow with 20% D-tagatose also indicate growth of liver tissue (Lina & de Bie, 1998a; Bär et al., 1999).

Groups of 25 adult male Sprague-Dawley rats received SDS diet containing 5% D-tagatose and 15% pregelatinized starch at the expense of 20% barley for 4 weeks, providing a dose of D-tagatose equal to 2.6-2.8 g/kg bw per day. The control group received SDS diet in which 20% barley was replaced by 20% pregelatinized potato starch. The study was performed according to GLP. After 4 weeks, 15 rats from each group were fasted for 24 h and then killed. The remaining 10 rats per group were not fasted prior to killing.

The mean body weights of animals given D-tagatose were slightly lower (3-4%) than those of controls. The absolute and relative weights of wet and dried liver were increased significantly in rats given D-tagatose that were killed without fasting when compared with non-fasted or fasted controls; however, the liver weights of non-fasted controls were greater than those of fasted controls. No difference in liver weights was seen between rats given D-tagatose and controls that were killed after a fast. The weights of the full and empty caecum were greater in non-fasted rats given D-tagatose than in non-fasted controls, but this effect was not seen in fasted rats; however, non-fasted control rats also had greater full and empty caecum weights than their controls (except for the relative weight of the empty caecum). Rats killed in a fasted condition showed no changes in the total protein, total lipid, or DNA content of the liver when compared with fasted controls. The glycogen contents of the livers of fasted treated and control rats were below the limit of detection; however, the livers of non-fasted rats given D-tagatose had significantly increased glycogen (expressed as g/100 g liver or as g/g liver) and DNA (as mg/mg liver), whereas the percentages of protein, total lipid, and residual moisture in the liver and DNA as mg/g liver and mg/g protein were unchanged. The amount of protein per g liver was increased slightly in non-fasted rats given D-tagatose, but non-fasted controls also showed increased mg protein/mg DNA and mg glycogen/mg DNA when compared with fasted controls. The authors interpreted these results to indicate that D-tagatose does not cause significant growth or hypertrophy of liver tissue. While the liver glycogen content was increased by D-tagatose, the larger amounts of glycogen were completely metabolized during a 24-h fasting period (Lina & de Bie, 1998b; Bär et al., 1999).

Groups of 20 adult male Sprague-Dawley rats received 0, 5, 10, or 20% D-tagatose in SDS diet at the expense of 20% barley, to give actual intakes of D-tagatose of 0, 2.8, 5.5, and 11 g/kg bw per day, for 29-31 days. The diets with 0, 5, and 10% D-tagatose were compensated by 20, 15, and 10% pregelatinized potato starch, respectively. The diets were fed for life. Twenty clinical parameters, including measures of uric acid, phospholipids, cholesterol, and triglycerides, were examined in 10 rats from each group at the end of treatment. The study was performed according to GLP.

No deaths or clinical signs were observed. The mean body weights were decreased significantly (although by < 10%) at all doses with a dose–response relationship throughout treatment in animals given 10 or 20% D-tagatose and during the last 2 weeks in those given 5% D-tagatose. The food consumption of animals given 10 or 20% D-tagatose was reduced slightly but statistically significantly during week 1. Food conversion efficiency was reduced in rats given 20% D-tagatose. Significantly increased plasma alanine and aspartate aminotransferase activity, albumin, albumin:globulin ratio, cholesterol, phospholipid, and inorganic phosphate concentrations were observed in rats at 20% D-tagatose; whereas other clinical chemical parameters were unchanged. The relative weights of the liver and empty caecum were increased significantly at all doses with a dose-response relationship. The relative weight of the filled caecum and the absolute weight of the caecum were increased significantly only in rats at 20% D-tagatose. No other tissues were weighed. Macroscopy showed gas in the colon of three to four rats in each treated group. The activities of palmityl coenzyme A oxidase and lauric acid 12-hydroxylase (indicators of peroxisome proliferation) in liver homogenates prepared from six animals per group were increased significantly in rats given 10 or 20% D-tagatose, but no electron microscopic evidence of peroxisome proliferation was found. According to the authors, the 1.4-2.1 times greater activity of the two enzymes in rats given 20% D-tagatose over control values demonstrates that D-tagatose is only a weak peroxisome proliferator. Determination of DNA synthesis in the liver by an immunohistochemical method for Ki-67 labelling indicated that D-tagatose did not induce proliferation of hepatocytes but caused significant hepatocellular hypertrophy, with a statistically significant, dose-related decrease in the mean number of nuclei per fixed area at 10 and 20% D-tagatose. Electron microscopy of both centrilobular and periportal zones of the livers of five animals in the control group and that at 20% D-tagatose revealed an increase in volume density of glycogen in those given D-tagatose but no changes in the endoplasmic reticulum, mitochondria, or the Golgi apparatus (Lina et al., 1998; Bär et al., 1999).

Groups of 20 male and 20 female Sprague-Dawley rats (the males weighing 160-220 g and the females 144-190 g) received 0, 5, 10, 15, or 20% D-tagatose in their diet for 90 days, equal to 2.3-6.7, 4.7-13, 7.3-20, and 9.9-26 g/kg bw per day for males and 2.8-6.3 6.0-12, 8.7-18, and 12-25 g/kg bw per day for females. An isocaloric control group of 20 male and 20 female animals received 20% of a 50/50 mixture of D-fructose and cellulose in their diet for 90 days. The animals were fasted overnight (12 h) before killing. The study was performed according to GLP and guidelines of the US Food and Drug Administration, which are identical to OECD Guideline No. 408 except for neurological parameters.

One animal each at 10, 15, and 20% D-tagatose died before the end of the study; the causes of death were considered not to be related to treatment. Ophthalmoscopy showed no abnormalities. Most of the animals given 15 and 20% D-tagatose had soft stools from days 1-3 but none by day 4. A slight but statistically significant decrease in mean body weights was observed in animals at 20% D-tagatose, in males on days 21-91 and in females on days 63, 84, and 91, with a 10% decrease at day 91, and in males at 15% on days 42-70 and day 84 in comparison with the normal controls. When compared with the isocaloric control group, a slight but statistically significant decrease in mean body weight (< 10%) was seen in males at 20% D-tagatose, but this effect occurred only on days 42-91 and was less pronounced than that seen in comparison with the normal control group. Males on 20% D-tagatose had a statistically significant decrease (< 10%) in food consumption during week 1 when compared with the normal control group; when compared with the isocaloric controls, all males treated with D-tagatose had statistically significant decreases in food consumption throughout treatment.

Males and females given 15 and 20% D-tagatose had statistically significantly decreased values for red blood cell parameters (haemoglobin, haematocrit, mean corpuscular volume, mean corpuscular haemoglobin, and mean corpuscular haemoglobin concentration (not in females given 15%)) when compared with the normal control group. Additionally, a statistically significant increase in erythrocyte count was seen in males given 20% D-tagatose, statistically significant decreases in activated partial thromboplastin times in females given 15 and 20%, and statistically significant increases in plasma fibrinogen in males at 15% and females at 20% D-tagatose. Males given 10% D-tagatose also had statistically significant decreases in mean corpuscular volume and mean corpuscular haemoglobin.

Statistically significant changes in clinical chemical parameters in comparison with the normal control group that were possibly related to treatment were dose-related decreases in alanine aminotransferase activity in males at 15% and females at 10, 15 and 20% D-tagatose, dose-related increases in cholesterol concentration in males and females at 15 and 20%, and dose-related increases in triglyceride concentration in males at 15 and 20%. The other statistically significant changes in clinical chemical parameters (decreases in serum alkaline phosphatase activity in males at 10 and 15%, increases in calcium concentration in males at 20%, increases in blood urea nitrogen in males at 10%, and increases in total protein and albumin concentrations in males at 15 and 20%) were considered not to be biologically significant. Uric acic concentrations were unchanged. The rectal temperatures of males in all treated groups were statistically significantly increased when compared with those of the normal control group, but the body temperatures were within normal limits and were therefore considered to be not biologically significant.

The absolute liver weights of males and females were increased statistically significantly in those at 15 and 20% D-tagatose and nonsignificantly in those at 10% when compared with the normal control group. The weights of the livers of males and females at 10, 15, and 20% relative to that of the body and that of the brain were statistically significantly increased, except for the liver:brain weight ratio of females at 10%. Statistically significant decreases in the absolute and relative weights of the epididymal fat pads were seen in animals at 15 and 20% and of the parametria at all doses; these changes were ascribed by the authors to shifts in fat stores and were considered not to be a toxic effect. Statistically significant changes in other organ weights (thyroid/parathyroid, thymus, ovaries, and mesenteric lymph node) were considered not biologically significant owing to the absence of dose-response relationships. Macroscopic examination showed thickening of the liver in 5/20 males at 20% D-tagatose, 4/19 males at 15%, and 1/19 at 10%. One of 20 males and 1/19 females at 20% showed discolouration of the liver. Microscopy revealed minimally enlarged centrilobular hepatocytes (hypertrophy) with an increased amount of a dense, eosinophilic cytoplasm in 7/20 males and 11/20 females at 20% and 4/20 males and 4/20 females at 15% D-tagatose. Examination of the livers of six animals of each sex in the control group and those at 20% D-tagatose by periodic acid-Schiff staining did not show an increased prevalence of centrilobular glycogen. No treatment-related microscopic changes were seen in animals at 5 and 10% D-tagatose.

The significant differences in body-weight gain and food consumption between animals at 20% D-tagatose and the isocaloric control group (isocaloric control males consumed more food than either those at 20% D-tagatose or normal controls and had a weight gain comparable to that of normal controls and animals at 5 and 10% D-tagatose; males at 20% D-tagatose did not have increased food consumption and had significantly reduced growth) suggest differences in the nutritional status of the isocaloric control group and that given 20% D-tagatose. The authors therefore considered that the isocaloric control group was not appropriate and suggested that the results for the D-tagatose groups should be compared only with those for the normal control group. The NOEL was 5% D-tagatose in the diet, equal to 2.3 g/kg bw per day (Trimmer et al., 1993; Kruger et al., 1999a).

A study was performed with maturing Sprague-Dawley rats in order to compare body-weight, fat, and lean mass gains between groups receiving D-tagatose and sucrose. Thirty retired male breeder rats were fed rat chow plus 15% D-tagatose ad libitum during a 14-day adaptation period. Thereafter, six rats were killed and the remaining rats were divided into four groups of six animals and received chow plus 15% D-tagatose ad libitum, chow plus 15% sucrose, chow plus 5% D-tagatose and 10% D-fructose, or chow plus 15% D-fructose for 90.5 days. The groups given 15% sucrose, 5% D-tagatose plus 10% D-fructose, and 15% D-fructose received the same amount of feed as that given 15% D-tagatose.

During the first two days of adaptation, 30/30 and 11/30 animals had soft or liquid faeces; however, on day 1 only one rat showed these effects, and the faeces of all animals were normal on day 4. The mean body-weight gains were 55 g for animals given D-tagatose, 81 g for those given D-tagatose and fructose, 90 g for those given D-fructose, and 130 g for those given sucrose, and the mean body-fat gains were 19, 34, 61, and 75 g, respectively. The absolute and relative liver weights of animals given D-tagatose were statistically significantly increased when compared with those of animals given sucrose or fructose. The absolute and relative liver weights of animals given D-tagatose plus fructose were not statistically significantly different from those given sucrose or fructose (Saunders, 1992).

In a study reported only in an abstract, two groups of nine diabetic obese rats (SHR/N-cp, sex not given) received a diet comprising 24% fructose, 10% glucose, 10% starch, 19% fat, 10% lactalbumin, 10% casein, 5.9% cellulose, 3.1 % AIN salt mix, and 1.0% vitamins to which 10% fructose or 10% D-tagatose was added, for 6 months. The group given fructose had marked polydipsia during the first 3 months, whereas the water intake of those given D-tagatose was normal. This group had increased food efficiency during the first 2 months, probably due to the absence of polydipsia, and decreased food efficiency during the last 2 months of the study when compared with the group given fructose (Szepesi et al., 1996).

Four groups of two pigs received a regular diet with or without 20% sucrose, 20% D-tagatose, or 10% sucrose and 20% D-tagatose according to the protocols shown in Table 2. Samples of liver were taken about 6 h after the last meal, with two samples for light microscopy and two for electron microscopy. Light microscopy showed normal architecture and structural elements. Only minimal mononuclear cell infiltration was present, except in one pig in group 3, but this was within normal limits for healthy young pigs, according to the authors. Electron microscopy revealed no treatment-related changes. Examination for the presence of altered intra-cytoplasmic smooth membranes, such as glycogen-associated membranous arrays, myelin figures, or circular or pleomorhic membranous formations, as seen in patients with hereditary fructose intolerance showed no such changes. Individual cell necrosis was seen in tisues from control animals. Focal membrane-bound radiolucent areas with decreased numbers of glycogen granules were present in tissues from animals given D-tagatose and controls (Mann, 1997).

2.2.3 Genotoxicity

The results of tests for the genotoxicity of D-tagatose are shown in Table 3. The tests were carried out before the latest OECD guidelines became valid but nearly completely meet those requirements and are considered to be acceptable.

2.2.4 Developmental toxicity

In a study to determine the range of doses for a study of developmental toxicity, six groups of five pregnant Sprague-Dawley rats received D-tagatose in deionized, distilled water orally by gavage at a dose of 0, 4, 8, 12, 16, or 20 g/kg bw per day on days 6-15 of gestation in two equal doses given at an approximately 4-h interval. The dams were killed on day 20 of gestation. The study was performed according to GLP and the 1982 guidelines of the US Food and Drug Administration and also met the guideline for segment II teratgenicity studies as described in the 1966 guidelines of the same Administration.

No deaths were observed. Unformed or watery stools were seen in dams at doses ³ 12 g/kg bw per day during days 6-11 of gestation; thereafter, the stools were normal. At doses ³ 8 g/kg bw per day, a dose-related decrease in maternal food consumption was seen during treatment. No effect was observed on pregnancy rate, body weight during gestation, or uterine implantation. The absolute and relative weights of the livers of the dams were normal, and no macroscopic changes were seen. In the fetuses, no effect on body weights or sex distribution was observed, and no external abnormalities were seen. Skeletal and visceral examinations of the fetuses were not performed (Schroeder, 1994a).

Groups of 20-24 pregnant Sprague-Dawley rats received D-tagatose in deionized, distilled water orally by gavage at a dose of 4, 12, or 20 g/kg bw per day on days 6-15 of gestation in two equal doses at an approximately 4-h interval. The dams were killed on day 20 of gestation. The study was performed according to GLP and the 1982 guidelines of the US Food and Drug Administration and also met the guideline for segment II teratgenicity studies as described in the 1966 guidelines of the same Administration.

No deaths occurred. Most dams at 12 and 20 g/kg bw per day had unformed or watery stools during treatment on days 6-8 of gestation, and some animals showed this effect throughout treatment. Dams at 20 g/kg bw per day showed decreased weight on days 6-9 of gestation but then gained weight, resulting in a weight gain comparable to that of controls on days 6-16. Dams at 12 and 20 g/kg bw per day also had reduced food consumption during days 6-9 of gestation. The absolute and relative (to the corrected day-20 weight) weights of the liver of dams at 12 and 20 g/kg bw per day were statistically significantly increased (except for the absolute weight at 12 g/kg bw per day), with a dose-response relationship. Histological examination of the livers showed no abnormalities. No effects were seen on other maternal parameters (numbers of corpora lutea and uterine implantation sites, preimplantation loss, number of resorptions, resorption:implant ratio, incidence of resorptions, number of viable fetuses) or on fetal parameters (body weight, liver weight, sex ratio, external, visceral, and skeletal malformations). The NOEL for teratogenic or embryotoxic effects was 20 g/kg bw per day, and that for maternal effects was 4 g/kg bw per day (Schroeder, 1994b; Kruger et al., 1999c).

2.2.5 Special studies

D-Fructose and D-tagatose stimulated the glucokinase-catalysed reaction of glucose to glucose-6-phosphate in isolated rat hepatocytes to the same extent in what was ascribed either to an antagonizing effect of the metabolite fructose-1-phosphate or D-tagatose-1-phosphate on the binding of glucokinase-regulatory protein to glucokinase (Van Schaftingen & Van der Cammen, 1989; Detheux et al., 1991; Van Schaftingen et al., 1997) or on translocation of glucokinase from its magnesium-dependent binding site (Agius, 1994). Glucose-6-phosphate is known to stimulate glycogen formation by activation of glycogen synthetase, which produces glycogen from normal dietary precursors of glucose. Fructose was found to stimulate glycogen synthesis by promoting the glucokinase-catalysed phosphorylation of glucose to glucose-6-phosphate in the liver, mediated by fructose-1-phosphate. The increased glycogen formation by D-tagatose was proposed to be caused by a corresponding mechanism (Bär, 1999b). In vivo, D-tagatose is more efficient than fructose in forming glycogen, owing to slower elimination of D-tagatose-1-phosphate than of fructose-1-phosphate by aldolase B in the liver (Rognstad, 1975, 1982; Bär, 1999a).

It has been postulated that the increase in normal liver mass seen in fasted rats fed 10 or 20% D-tagatose is triggered by increased postprandial storage of liver glycogen, resulting from simultaneous feeding of D-tagatose and glucose equivalents. In order to investigate this hypothesis, the effects of separate and simultaneous administration of D-tagatose and glycogen precursors were investigated on liver weight and glycogen concentration in Wistar and Sprague-Dawley rats during periods of inactivity (dark period) and activity (light period). Four groups of 20 male Wistar rats received the following diets for 7 days: 6 g/day of polycose (glucose polymer) in 18 g/day of a restricted, glucose-free diet at the start of the dark period and carrier (2 g protein in water) by gavage at the end of the light period; D-tagatose in 18 g/day of a restricted, glucose-free diet (mean D-tagatose intake, 5.8 g/kg bw per day) at the start of the dark period and polycose by gavage at the end of the light period; D-tagatose and polycose in 18 g/day of a restricted, glucose-free diet (mean D-tagatose intake, 7.2 g/kg bw per day) at the start of the dark period and carrier by gavage at the end of the light period; or D-tagatose and polycose in two equal portions of 9 g/day of glucose-free diet (mean D-tagatose intake, 6.7 g/kg bw per day) at the start and end of the dark period and carrier by gavage at the end of the light period. After 7 days, 10 rats/group were killed at the end of the dark period and 10 rats/group at the end of the light period. Treatment continued until the animals were killed.

Weight loss was seen in all groups, probably due to the restricted feeding and low energy intake. Simultaneous administration of D-tagatose and glucose was assumed to result in glycogen deposition and to increase the normal liver mass, as indicated by the significantly lower glycogen deposition and weight of the liver in rats given D-tagatose at 5.8 g/kg bw per day followed by polycose and killed at the end of the light period when compared with those in rats given D-tagatose at 7.2 g/kg bw per day with polycose and killed at the end of the dark period. The livers of rats in the latter group that were killed at the end of the light period contained no glycogen but weighed significantly more than those of controls given polycose only and killed at the same time, probably reflecting early liver growth in response to D-tagatose-induced glycogen deposition. Rats given D-tagatose and polycose in two equal portions of 9 g/day of glucose-free diet (mean D-tagatose intake, 6.7 g/kg bw per day) at the start and end of the dark period and carrier by gavage at the end of the light period and then killed at the end of the dark period showed slightly increased liver weights and unaffected glycogen levels when compared with controls. Apparently, division of the feed resulted in more regular intake without glycogen peaks and with no triggering of glycogen deposition and liver growth (Lina & de Bie, 2000a).

In a similar study, three groups of 18 adult male Sprague-Dawley rats received the following diets for 5 days: 6 g/day of polycose in 16 g/day of a restricted, glucose-free diet at the start of the dark period and carrier by gavage at the end of the light period; D-tagatose in 16 g/day of a restricted, glucose-free diet (mean D-tagatose intake, 7.7 g/kg bw per day) at the start of the dark period and polycose plus carrier (2 g protein in water) by gavage at the end of the light period; or D-tagatose and polycose in 16 g/day of a restricted, glucose-free diet (mean D-tagatose intake, 8.8 g/kg bw per day) at the start of the dark period and carrier by gavage at the end of the light period. After 5 days on the diets, groups of six rats were killed at 06:00 h during the dark feeding period, at 12:00 h before gavage, and at 18:00 h after the last gavage.

Weight loss was seen in all groups, due to the restricted feeding. Rats receiving D-tagatose and polycose and killed at 6:00 h had increased glycogen concentrations and heavier livers than controls. In rats killed at 12:00 and 18:00 h, when no food was available, the glycogen concentrations in the livers of rats receiving D-tagatose and polycose and of controls were decreased, although those of treated rats remained slightly higher than those of controls. The glycogen levels and liver weights of rats given D-tagatose and then polycose and killed at 6:00 and 12:00 h were lower than those of rats given D-tagatose and polycose simultaneously and also lower than those of controls. At 18:00 h, rats given D-tagatose and then polycose had liver weights comparable to those of controls, whereas the glycogen contents were greater than those of rats given the two substances simultaneously and also of controls, probably due to the administration of glucose-free diet during the dark period and polycose in the light period.

Three additional groups of 12 rats were intended to be used for various examinations of the liver after 4 weeks on the diets described above. However, during week 2 of treatment, the condition of the rats declined and gavage became increasingly difficult. The test was terminated on day 10 (Lina & de Bie, 2000b).

Owing to the fact that L-sorbose, which is also a stereoisomer of D-fructose, caused haemolysis in dog erythrocytes, this phenomenon was also examined with D-tagatose. Washed dog erythrocytes were incubated in Hank’s balanced salt solution containing 5.6 mmol/L glucose with D-tagatose at 0.6, 6, or 60 mmol/L, L-sorbose at 0.6, 6, or 60 mmol/L, or glucose added at 0, 0.4, or 54.4 mmol/L for 24 h at 34 °C. No haemolysis was seen with D-tagatose or glucose, whereas incubation with 6 or 60 mmol/L L-sorbose resulted in significant haemolysis (Bär & Leeman, 1999).

D-Tagatose had a cytoprotective effect in murine hepatocytes against oxidative stress induced by nitrofurantoin or cocaine, iron overload induced by ferric nitrilotriacetate, luminol-enhanced chemiluminescence, and apoptosis induced by the bile salt glycogeno-deoxycholate (Valeri et al., 1997; Zeid et al., 1997; Paterna et al., 1998). D-Tagatose did not protect against lethal cell injury induced by tert-butyl hydroperoxide, a prooxidant which acts by hydroxyl radical-independent mechanisms and which is partitioned in the lipid bilayer (Paterna et al., 1998).

The possibility of a non-enzymatic reaction of a D-tagatose with proteins (Maillard reaction) has been investigated. This reaction can occur, for example, in patients with poorly controlled diabetes during hyperglycaemic episodes and results in increased formation of glycosylated haemoglobin. In a study of reactivity with amino groups of haemoglobin, D-tagatose was one of the least reactive sugars. Reactivity appeared to be dependent on the extent to which the sugar existed in open (carbonyl) rather than ring (hemiacetal or hemiketal) form (Bunn & Higgins, 1981).

In rat skeletal muscle, D-tagatose showed low reactivity with myofibrillar proteins, and the reactivity was again dependent on the percentage of the sugar present in the open-chain form (Syrovy, 1994).

Several studies have been performed to examine the intestinal tolerance and the possible hyperuricaemic effect of D-tagatose in humans. The hyperuricaemic effect of D-tagatose was investigated because the degradation of D-tagatose-1-phosphate, the initial metabolite of D-tagatose, appears to be slower than that of D-fructose-1-phosphate, the initial metabolite of fructose, and because D-fructose increased uric acid production by accelerating degradation of purine nucleotides, probably due to hepatocellular depletion of inorganic phosphorus by accumulation of fructose-1-phosphate.

In a double-blind cross-over study, eight fasted men received a single oral dose of 30 g of D-tagatose or D-fructose in 400 ml of water or only water 4.5 h before lunch. The three solutions were administered at intervals of 5-33 days. The concentrations of uric acid in serum and urine were increased to a greater extent after ingestion of D-tagatose than with fructose. The peak serum concentration of uric acid (7.2 mg/100 ml; upper limit of normal range, 7.6 mg/100 ml) and the area under the curve of concentration-time (AUC) at 4 h of serum uric acid were significantly higher with D-tagatose than with D-fructose or water. The serum inorganic phosphorus concentration was lower 50 min after D-tagatose than with water, whereas no effect was seen with D-fructose. D-Tagatose increased the concentration of glucagon-like peptide 1 in plasma, but those of gastric inhibitory peptide and cholecystokinin were not affected. Although D-fructose had marked thermogenic and lactacidaemic effects, such effects were trivial after ingestion of D-tagatose. Serum glucose, insulin, and glucagon concentrations were not affected by ingestion of D-tagatose or fructose (Buemann et al., 1999a).

Five healthy fasting men received a single oral dose of 30 g of D-tagatose or 30 g of D-fructose dissolved in 400 ml of water in a double-blind cross-over study with an interval of > 6 days between experiments. Other food was not allowed for 4 h after dosing. The serum uric acid concentration was increased by 17% 50 min after ingestion of D-tagatose and had not reached the pre-treatment level by the end of the study 230 min after dosing. Although renal fractional extraction of uric acid decreased by 12%, this could not explain the acute hyperuricaemic effect (Buemann et al., 1999b,c).

Four healthy women and four healthy men and four women and four men with type 2 diabetes received a single dose of 75 g of glucose followed by a 3-h glucose tolerance test on day 1. Thereafter, all subjects were adapted to D-tagatose by ingesting single doses of 5, 10, and 25 g/day sequentially on days 2-4. On day 5, each subject was given a single oral dose of 75 g of D-tagatose followed by a 3-h tolerance test. The tolerance testing was conducted at 08:00 h after a 12-h fast. Blood samples were collected at 0, 30, 60, 120, and 180 min on days 1 and 5 after the doses of glucose and D-tagatose, respectively, and magnesium, phosphorus, and uric acid concentrations in plasma were determined. D-Tagatose increased the plasma uric acid concentration in all subjects, with a peak of about 1.5 mg/dL at 60 min, but the increases were not statistically significant and were within normal limits, except in the healthy men, who had mean uric acid concentrations above the upper limit of the normal range of 3.5-8.5 mg/dL. The concentrations had decreased in all subjects by 3 h but did not regain the pre-treatment level in 11/16 subjects. The AUC for uric acid in all subjects was statistically significantly increased in the D-tagatose tolerance test when compared with that in the glucose tolerance test. The highest values for the uric acid AUC in the D-tagatose tolerance test were found in healthy men and then in men with diabetes. The magnesium and phosphorus concentrations were not changed statistically or biologically significantly (Saunders et al., 1999b).

In a screening test, 73 healthy male volunteers received a single oral dose of 29-30 g of D-tagatose in firm cake in the afternoon. Nausea was reported by 15% of subjects and diarrhoea by 32% (Buemann et al., 1999c).

In a double-blind cross-over study with a 4-day interval between treatments, 23 women and 11 men received a single oral dose of 29 g of D-tagatose or sucrose in a continental breakfast. Gastrointestinal side-effects were reported on a five-level scale for 72 h after dosing. The scores for rumbling in the stomach and gut, distension, nausea, flatulence, and diarrhoea were significantly higher than after ingestion of sucrose, but only the score for distension remained high for more than 24 h. One case of vomiting after D-tagatose intake was observed (Buemann et al., 1999d).

Fifty healthy men and women aged 18-24 received 40 g of plain chocolate with 20 g of sucrose, lactitol, or D-tagatose in two equal portions between 13:00 and 16:00 h in a double-blind, controlled cross-over study with an interval of 1 week between treatments. The occurrence of colic, flatulence, borborygmus, and bloating was significantly increased after consumption of chocolate with lactitol or D-tagatose as compared with sucrose. Significantly increased incidences of nausea, appetite loss, and thirst were seen after consumption of D-tagatose as compared with sucrose and lactitol. The number of stools and their consistency were not affected (Lee & Storey, 1999).

In a double-blind cross-over study, eight volunteers received 30 g of D-tagatose or 30 g of sucrose per day at lunch for 15 days, with an interval of more than 2 months between the two treatments. Gastrointestinal symptoms, in particular nausea and flatulence, were reported more frequently and with a higher score after ingestion of D-tagatose than of sucrose. Adaptive gastrointestinal tolerance was not seen during the 15 days of treatment. Only two cases of severe symptoms were reported with D-tagatose, one of distension and one of loss of appetite. The values for body weight, fat-free mass, fat mass, and blood pressure were comparable with the two sugars, and no differences were seen between the values for day 1 and day 15. Increased hydrogen production was seen within 1.5 h after ingestion of D-tagatose, and the increases were similar on days 1 and 15. No significant changes were seen in plasma alkaline phosphatase or gamma-glutamine transferase activities or in uric acid, lipid, glucose, or insulin concentrations at the three sampling times on days 1, 7, and 15. Lower serum insulin and triglyceride concentrations and higher 24-h urinary uric acid concentrations (2.6 mmol/L with D-tagatose and 2.3 mmol/L with sucrose) were seen when the results for days 1, 7, and 15 were combined (Buemann et al., 1998, 1999c).

Two healthy men and two healthy women and two men and two women with type 2 diabetes received 25 g of D-tagatose orally three times a day (with each main meal), 7 days/week for 8 weeks. Two healthy men and two healthy women received 25 g sucrose three times a day for 8 weeks, whereas two men and two women with diabetes received no sugar supplementation for 8 weeks. Body weight, blood pressure, and plasma (fasted) magnesium, phosphorus, copper (only in eight healthy subjects), uric acid, cholesterol (total, high-density and low-density lipoproteins), triglycerides, glycosylated haemoglobin, glucose, insulin, alanine and aspartate aminotransferase and alkaline phosphatase activity (only after 6 and 8 weeks), and total bilirubin (only after 6 and 8 weeks) were determined on day 0 and every 2 weeks thereafter in all subjects. Flatulence was reported by 7/8 subjects and diarrhoea by 6/8 subjects given D-tagatose during the 8-week treatment period. No treatment-related increase in fasting plasma uric acid concentration or other parameters was observed (Saunders et al., 1999b).

Six healthy subjects received 75 g of D-tagatose or sucrose daily for 8 weeks. No significant changes were seen in either group in fasting blood glucose or insulin concentration, glycohaemoglobin, blood pressure, weight, or lipids. In four people with type 2 diabetes who received 75 g of D-tagatose daily for 8 weeks, a substantial decrease in glycohaemoglobin but no changes in fasting blood glucose concentration, blood pressure, lipids, or weight were seen (Donner et al., 1996).

Four healthy men and four healthy women and four men and four women with type 2 diabetes received D-tagatose orally at doses of 5, 10, and 25 g/day sequentially during an adaptation period of 3 days. Thereafter, they received a single oral dose of 75 g of glucose followed by a 3-h glucose tolerance test, a single oral dose of 75 g of D-tagatose followed by a 3-h tolerance test, or 75 g of D-tagatose 30 min before a 75-g glucose load and a 3-h glucose tolerance test. Between treatments, all subjects received D-tagatose at 10 g/day. Ingestion of D-tagatose alone did not increase the blood glucose or insulin concentration. When D-tagatose was given 30 min before the glucose load, the blood glucose concentration increased less than after a glucose load only; the effect was statistically significant in the persons with diabetes 60, 120, and 180 min after the glucose load and not significant in the healthy subjects 30 and 60 min after the glucose load. The AUC for glucose was significantly smaller in the patients who had been pretreated with D-tagatose. Pretreatment with D-tagatose tended to attenuate the increase in insulin concentra-tions in the healthy subjects after the glucose load. The AUC for insulin tended to be smaller after pretreatment with D-tagatose in healthy subjects but not in those with diabetes.

In 10 persons with type 2 diabetes given 0, 10, 15, 20, or 30 g of D-tagatose orally before a 75-g oral dose of glucose, a dose-related attenuation of the rise in plasma glucose concentration was observed.

Five persons with type 2 diabetes underwent a tolerance test with 75 g sucrose before or 30 min after a single oral dose of 75 g D-tagatose. Pretreatment with D-tagatose attenuated the rise in plasma glucose concentration after an oral dose of sucrose. No significant difference in the AUC for glucose or insulin was seen after ingestion of D-tagatose.

At the end of treatment, all 16 volunteers in the first test received increasing doses of D-tagatose with each meal (5 g with each meal on day 1, increased by 5 g per meal on each subsequent day) to establish the dose at which gastrointestinal effects first appeared. Administration of 75 g of D-tagatose led to diarrhoea, nausea, and/or flatulence in all subjects. At doses of 10-30 g, only 3 of the 10 subjects with diabetes had mild gastrointestinal symptoms (Donner et al., 1996, 1999).

Four men and four women with type 2 diabetes received 45 g of D-tagatose per day (15 g per main meal) orally for 12 months. Two months before treatment, on day 0, and at 2-month intervals, body weight was recorded, blood pressure was measured, and the following end-points in serum were analysed: albumin, alkaline phosphatase, amylase, bilirubin, calcium, cholesterol (total, high- and low-density lipoproteins), chloride, carbon dioxide, creatinine, glucose, insulin, glycohaemoglobin, potassium, magnesium, sodium, phosphorus, total protein, alanine and aspartate aminotransferases, triglycerides, blood urea nitrogen, and uric acid. Statistically significant decreases in body weight (3.5 kg) and in glycohaemoglobin were seen, starting after 4 months and continuing throughout the study. No changes in uric acid concentrations were seen in men or women, although the women had higher concentrations at the start of treatment which remained higher during treatment. The amount of carbon dioxide decreased up to month 6 and then returned to the baseline value (Makris, 1999). According to Bär (1999a), five subjects in this study reported increased flatulence during the first 2 weeks of treatment, and one women reported transient diarrhoea during the first few days.

3.1.1 Simulation of intake in Australia and the European Union

The predicted intake of D-tagatose in Australia was based on data on food consumption derived from the 1995 National Nutrition Survey for the population aged 2 years and over, comprising 13 858 individual dietary records and single 24-h recall. It was assumed that D-tagatose occurred in the food categories named by the manufacturer (including chewing-gum in the confectionery category) at the maximum nominal concentrations. An additional intake estimate was undertaken, assuming that D-tagatose occurred only in low-energy edible ices, confectionery, and water-based flavoured drinks and in all breakfast cereals.

The estimated mean intakes of D-tagatose, assuming use in the entire category of foods, were 7.3 g/day for all respondents and 9.7 g/day for consumers only. The estimated mean intakes of D-tagatose, assuming use only in low-energy foods, were 3.4 g/day for all respondents and 7.4 g/day for consumers only (Table 4). The intake of D-tagatose by consumers in the 95th percentile of intake was predicted to be 26 g/day assuming that D-tagatose is present in all the named foods and 18 g/day assuming restriction to low-energy foods (Table 4). It should be noted that 24-h dietary recall tends to result in overestimates of habitual intake.

    See Also:
       Toxicological Abbreviations