Re‐evaluation of fatty acids (E 570) as a food additive

Human studies

Schwabe et al. (1964) studied the absorption and oxidation of orally ingested or intravenous (i.v.) injected ¹⁴C‐carboxyl‐labelled caprylic acid, in 12 normal volunteers (medical students, no further details). The absorption and oxidation was determined by monitoring the expired ¹⁴CO₂ for 50 min after administration. In both trials, 2–3 μCi of the test item was given to each volunteer with a time interval of 3 days between oral or i.v. treatment. There was no significant difference in the mean cumulative recovery of ¹⁴CO₂ after oral or intravenous administration of labelled caprylic acid: 15.4% vs 15.7% of the applied radioactivity, respectively. A slight increase in the delay time (the time period between application and first measurement of radioactivity) was reported after oral administration (3–6 min) vs i.v. injection (1–2 min). In both trials, the maximal radioactivity in expired air was measured 15–30 min after administration. In further experiments, it was shown that co‐application of glucose, orally or by co‐infusion of 1 g unlabelled caprylic acid did not inhibit caprylic acid oxidation based on expired ¹⁴CO₂ measurements. The authors concluded that caprylic acid is readily absorbed from the gastrointestinal tract and rapidly oxidised.

In the studies by Chen et al. (2003) and by Sanaka et al. (2006), the absorption of ¹³C‐labelled caprylic acid was evaluated using the gastric emptying breath test. According to the authors, this test allowed indirect assessment of gastric emptying by monitoring exhaled ¹³CO₂. ¹³C‐Caprylic acid was absorbed only after it was emptied from the stomach, and then was oxidised to ¹³CO₂. The methodological relevance of the breath test was based on the assumption that intestinal absorption and hepatic oxidation progress rapidly. After an overnight fasting, breath samples were taken from six male healthy volunteers (30–39 years) who received, via an oesophago‐gastro‐duodenoscope, 100 mg ¹³C‐caprylate dissolved in 20 mL normal saline, in the descending portion of the duodenum. Maximal exhalation rate of ¹³CO₂ was found 15 min after application. Beyond 60 min post‐dose, ¹³CO₂ excretion diminished in an exponential manner, characterised by a half‐excretion time of 71.3 ± 12.7 min.

Animal studies

Blomstrand (1955) studied the intestinal absorption of [1‐¹⁴C]‐capric acid in four male rats (strain not given) with cannulated thoracic lymph ducts after gavage of olive oil containing 1.1% of the free [1‐¹⁴C]‐capric acid (no further information given). The lymph (in four experimental animals) and expired air (in three experimental animals) were collected for 24 h, after which the animals were sacrificed and the radioactivity of the whole gastrointestinal tract plus faeces was measured. The applied radioactivity was almost completely absorbed within 24 h from the gastrointestinal tract (96–100%) and 3–16% of the label was recovered in the lymph lipids from the thoracic duct. About 2% of the activity found in the lymph lipids was in the form of phospholipids but 98% in the form of triglycerides; only traces of free capric acid were detected. Large amounts of applied radioactivity (19% within 8 h and 46% within 21 h) occurred in the expired air in form of ¹⁴CO₂. Overall, free capric acid was nearly completely absorbed from the gastrointestinal tract and mainly metabolised and excreted in the form of exhaled CO₂; minor amounts were transported via the thoracic‐duct lymph and incorporated into triglycerides.

Borgström (1955) studied the transport form of ¹⁴C‐lipid in blood samples taken from the portal vein and from the inferior vena cava blood. Samples were collected 1.5 h after administration of 1 mL olive oil containing 28% free ¹⁴C‐capric acid administered via gavage to two adult rats (no further details provided). Higher radioactivity of blood lipids was detected in the portal vein (2.3‐ or 4‐fold increase compared to vena cava), mostly in the form of free fatty acids.

Vallot et al. (1985) studied, by perfusion of isolated duodenojejunal loops in male Wistar rats (n = 4), the modification of both the portal vein and intestinal lymph absorption of capric acid in the presence of a monoglyceride (monopalmitin) and long‐chain fatty acids (oleic‐ or palmitic acid). In the experiments, 10–15 μCI (52 mCi/mmol) of 1‐¹⁴C‐capric acid was added to each lipid emulsion. Blood samples were collected from the cannulated portal vein at 5 min intervals for 1 h after infusion of 90 μmol of labelled capric acid. During this time period, 80% of the applied radioactivity disappeared from the intestinal loop and 11% was recovered in the mucosa at termination (after 1 h). About 35% of the capric acid was transported directly into the bloodstream mainly in form of free fatty acids. Capric acid oxidation in the mucosa was evident since ¹⁴CO₂ was detected in portal venous blood. The authors concluded that labelled capric acid was rapidly and mainly absorbed via the portal route. When, in further experiments, labelled capric acid was infused into the loop of intestinal lymph fistulated rats (n = 3), a very low percentage of radioactivity (< 1%) was recovered in the lymph during a sampling period of 6 h. The intestinal absorption and mucosal oxidation of capric acid was modulated by the co‐application of palmitic‐ or oleic acid.

Animal studies

Göransson (1965) administered i.v. 0.5 mL rat serum containing 0.1 μeq [1‐¹⁴C]‐labelled lauric acid together with 0.1 μeq [9,10‐³H]‐labelled palmitic acid to fasted and refed rats (no further details given). The disappearance of the two different labels (¹⁴C vs ³H) was measured 1–5 min after injection. Lauric acid disappeared more rapidly from blood than palmitic acid. Data on tissue distribution of the two labels indicated that lauric acid was more rapidly oxidised than palmitic acid. Similar results were reported in the re‐fasted rats.

Animal studies

Göransson (1965) administered i.v. 0.5 mL rat serum containing 0.1 μeq [1‐¹⁴C]‐labelled myristic acid together with 0.1 μeq [9,10‐³H]‐labelled palmitic acid to fasted and refed rats (no further details given). The disappearance of the two different labels (¹⁴C vs ³H) was measured 1–5 min after injection. Myristic acid disappeared more rapidly from blood than palmitic acid. Data on tissue distribution of the two labels indicated that myristic acid was more rapidly oxidised than palmitic acid. Similar results were reported in the non‐fasted rats.

Rioux et al. (2000) compared the metabolism of myristic and palmitic acids in cultured hepatocytes obtained from male Sprague–Dawley rats. [1‐¹⁴C]‐Labelled fatty acids were incubated for 12 h with 24‐h cultured hepatocytes. Myristic acid was taken up more rapidly than palmitic acid (about 87% and about 69%, respectively; p < 0.05). After 4 h of incubation, 13% of the initial myristic acid label was still in the medium while after 12 h of incubation there was almost no myristic acid left in the medium (0.9%). Incorporation into cellular lipids, however, was similar after the same time (33.4 ± 2.8% and 34.9 ± 9.3%, respectively, of initial radioactivity). In the early phase of incubation (30 min), myristic acid was more rapidly incorporated into cellular triglycerides than was palmitic acid (7.4 ± 0.9% and 3.6 ± 1.9% of initial radioactivity respectively). However, after 12 h incubation, the radioactivity of cellular triglycerides, cellular phospholipids, and secreted triglycerides was significantly higher with palmitic acid as precursor. Myristic acid oxidation was significantly higher than that of palmitic acid, with 14.9 ± 2.2% and 2.3 ± 0.6% of the initial radioactivity respectively, being incorporated into β‐oxidation products after 4 h. Myristic acid was also elongated to palmitic acid to a greater degree (12.2 ± 0.8% of initial radioactivity after 12 h) than palmitic acid was to stearic acid (5.1 ± 1.3% of initial radioactivity after 12 h). The combination of elongation and β‐oxidation resulted in the rapid disappearance of C14:0 in hepatocytes whereas C16:0 was esterified to form glycerolipids.

Animal studies

Coniglio and Cate (1959) administered 100 mg [1‐¹⁴C]‐palmitic acid in 1 mL of olive oil to fed male Sprague–Dawley rats (tissue from two to four animals were pooled per time point) via a stomach tube (no further details given). The animals were sacrificed 3, 6 or 24 h after administration. The intestine was washed out with water and lipids of the intestinal contents, intestinal tissue and the liver were extracted and analysed for distribution of the ¹⁴C in various types of lipids (phospholipids, glycerides and fatty acids). The total activity in the intestinal contents was predominantly as free fatty acids at both 3 and 6 h after administration. However, up to 18% of the radioactivity was found in the triglyceride fraction at these times. In the intestinal tissue, radioactivity was associated largely with triglycerides, but at 6 and 24 h, this fraction contained decreasing amounts of radiolabel and there was a progressive increase in radioactivity associated with phospholipids. In the liver, however, the activity was greater in phospholipids than in triglycerides measured 3 h after application; minor amounts of label were also found in mono‐ and diglycerides as well as cholesterol esters.

Laurell (1959) studied the distribution of radiolabelled free palmitic acid in non‐fasted or fasted male albino rats (n = 3 each; strain not given) after i.v. injection of 8–10 μCi of ¹⁴C‐ labelled palmitic acid in 1 mL 10% albumin or in 1 mL rat serum. Twenty minutes after injection, rats were sacrificed and liver, muscle and adipose tissues were removed and lipids extracted and analysed. Radioactivity in other organs and carcass was measured only in one rat. In non‐starved rats, the total recovery ranged between 83% and 94%. Most activity was found in the liver: 40–43% of applied radioactivity (20–28% in triglycerides, 15–19% in phospholipids and only traces of free palmitic acid), 6–8% in adipose tissues but less than 1% in 2 g muscle (mainly as triglycerides in both tissues). Only 8–11% of applied radioactivity was detected in all other internal organs (mainly digestive tract) but 5–8% in the skin and 21–28% in carcass. Similar results were obtained in starved rats; however, the total recovery was reduced to 50–59% which might be related to rapid metabolism and exhalation of radiolabel (not measured). Overall, most of the radioactivity was found in the liver after i.v. injection of labelled free palmitic acid.

Olivecrona (1962) administered 5 μCi of ¹⁴C‐labelled palmitic acid (bound to albumin) via i.v. injection to male Sprague–Dawley rats (number of animals not stated). Data showed that 30% of the label was taken up by the liver within 5 min where almost 97% of the label was present in the esterified form. More label appeared in the hepatic neutral fat fatty acids than in the phospholipid fraction and it disappeared more rapidly from neutral fat, suggesting rapid turnover. Similar relations were found for plasma neutral fat fatty acids versus phospholipids. One h after injection 3% of the applied radioactivity was detected in adipose tissue. Eighty minutes after application the levels of ¹⁴C in plasma and liver lipids were similar suggesting equilibration of the two pools. However, the carcass lipid pool was still lower than plasma and liver pool even after 24 h.

Clandinin et al. (1988), administered [1‐¹³C]‐palmitic acid and [16‐¹³C]‐palmitic acid (in capsule form) with a breakfast meal to four male healthy volunteers (22–27 years old) at a dose 0.5–0.75 mg ¹³C/kg bw. The ¹³CO₂ was measured in exhaled air. Chain shortening of palmitic acid was examined by comparing oxidation rates of [1‐¹³C]‐palmitic acid vs [16‐¹³C]‐palmitic acid. The whole‐body rate of oxidation of [1‐¹³C]‐palmitic acid was significantly greater than that observed for [16‐¹³C]‐palmitic acid. These results suggested that up to 34% of dietary palmitic acid consumed may be subjected to chain shortening.

Animal studies

Carroll and Richards (1958) studied the digestibility of stearic acid in rats (Sprague–Dawley and Wistar rats; only limited information was provided on methods and animals). The animals were fed a diet containing 10% unlabelled stearic acid (equivalent to 11,800 mg/kg bw per day) for 16 days. The digestibility of stearic acid was measured by analysis of ingested fat and the fat excreted via faeces. The authors calculated a digestibility of 24%. For comparison, a digestibility of 40% was obtained in similar trials with palmitic acid and 73% with oleic acid.

De Leo and Foti (1961) administered 200 μCi 1‐¹⁴C labelled stearic acid in 1 mL olive oil via gavage to four male Wistar rats. The results showed the formation of labelled cholesterol, which was widely distributed in the organism within 24 h after administration; the highest activity of cholesterol was noted in liver and serum.

Animal studies

Bergström et al. (1954) studied the intestinal absorption of oleic acid in five male rats (strain not given) with cannulated thoracic lymph ducts, after gavage of 0.5 mL/rat of [1‐¹⁴C]‐oleic acid (no further data). The lymph and expired air were collected for 24 h, after which the animals were sacrificed and radioactivity in the whole gastrointestinal tract plus faeces were measured. A mean of 78% of the fed activity (range: 69–92%) was absorbed in 24 h and 50% of the applied radioactivity was recovered in the lymph lipids from the thoracic duct. About 2% of the activity found in the lymph lipids was in the form of phospholipids and 98% in the form of glycerides. Only 4% of applied activity was exhaled via labelled CO₂ in cannulated rats but 17% in a rat without a thoracic duct fistula. The authors concluded that oleic acid was transported via the thoracic lymph duct and incorporated mainly into triglycerides.

Carroll and Richards (1958) measured the digestibility of oleic acid by analysis of ingested fat and of fat excreted via faeces in rats (Sprague–Dawley and Wistar rats; only limited information was provided on methods and animals) fed a diet containing 10% (equivalent to 11,800 mg/kg bw per day) unlabelled oleic acid for 16 days. The authors calculated a digestibility of 73%, but lower values were obtained in similar trials with saturated fatty acids: 40% digestibility for palmitic acid and only 24% for stearic acid.

In a similar experiment by Gallagher et al. (1965) with male Sprague–Dawley rats, the thoracic lymph duct was cannulated prior to application via a gastrostomy tube of 0.1 mL [U‐¹⁴C]‐oleic acid (purity 99.5%) diluted with unlabelled oleic acid (no further details given). The lymph was collected for 12 h after feeding the labelled oleic acid. Thereafter, the animals were sacrificed and the radioactivity in the whole gastrointestinal tract plus faeces was determined. After application into the stomach, the mean absorption of applied radioactivity from the gastrointestinal tract was 82% (n = 12) and radioactivity detected in the lymph lipid within 12 h after application was 69% (mainly in lymph obtained the first 4 h). Only a small amount of label in the lymph lipid was identified as free oleic acid (4%) but essentially in triglycerides. After application into the duodenum (n = 22) the absorption of applied radioactivity decreased to 68%.

Kern and Borgström (1965) studied the absorption of [9,10‐³H]‐labelled oleic acid in albino rats (16 animals; weight, 150–200 g). A suspension (3 mL) of the test item containing skimmed milk, glucose, unlabelled oleic acid or labelled oleic acid (3 μCi) was administered via gavage, and the animals were sacrificed 2, 4, 6 or 9 h after application. The gastrointestinal tract along with its contents were removed and analysed for radioactivity. In the gastrointestinal tract, after the given time intervals, 72%, 62%, 49% or 29% respectively, of the applied radioactivity was detected. Significantly lower values were measured in similar trials after application of a suspension containing the bile salt sodium taurodesoxycholate (120 μM).

Hamilton (1971) studied the absorption of oleic acid in vivo by the rat jejunum, measuring the uptake, incorporation into triglyceride and transport out of the intestine by tied loops of jejunum and, in vitro, by measuring the incorporation into triglyceride by jejuna slices. For that purpose, 0.25 mL of lipid solution in 15 mM sodium taurocholate containing 2.4 mg/mL of [¹⁴C]‐oleic acid (purity 96%) at pH 5.8 or 7.2 (emulsion at pH 5.8 or micellar solution at pH 7.2) were placed in two different rat jejunal loops of male Sprague–Dawley rats. The loops were washed‐out at ‘zero’ time and at 10 or 60 s, 5 or 30 min. For each trial, 4–10 rats were used. There was a rapid initial uptake of labelled oleic acid into the mucosa, with a slower incorporation into triglyceride and disappearance out of the loop. The absorption of radioactivity at pH 5.8 and pH 7.2 were similar at all time points. Absorption was related only to the total oleic acid concentration, regardless of physical form (i.e. emulsion or micellar solution) in which the oleic acid was present.

Cunningham and Lawrence (1977) administered by gavage 15 μCi [³H]‐labelled oleic acid in 1 mL olive oil to four male Wistar rats (100 g bw). After 24 h, 93% of the applied radioactivity was absorbed from the gastrointestinal tract, only 7% of the radioactivity was detected in the faeces (minor amounts were present in bile from separately treated animals collected over 3 days), 4% of the radioactivity was excreted via the urine.

Hollander et al. (1984) studied the intestinal absorption of oleic acid [1‐¹⁴C]‐labelled, purity > 99%, activity 56 Ci/mol) at low intraluminal concentrations and the influence of luminal factors on its absorption, in male Sprague–Dawley rats. The labelled test solution was applied into the small bowel via an inflow tube at 1 cm distal to the common bile duct; a glass tube, 50 cm distal to the infusion site, was used to cannulate the intestine for sampling the outflow for a period of 1 h in 10 min aliquots followed by analysis of radioactivity. The labelled oleic acid was diluted with unlabelled oleic acid (purity 99%). Concentrations of 30–2,500 μM oleic acid in a 10 mM sodium taurocholate micellar solution (pH 6.5) were tested (n ≥ 3 per trial). Absorption of oleic acid increased as the pH was decreased and taurocholate concentrations were increased or as the thickness and resistance of the unstirred water layer were diminished or following addition of lysolecithin. The additions of Tween‐80 (polysorbate 80), lecithin, linoleic‐, linolenic‐ or arachidonic acid to the perfusate decreased the rate of absorption of oleic acid. The authors concluded that oleic acid absorption displayed apparent saturation kinetics which was due to unstirred layer effects, limited aqueous solubility of oleic acid and possible saturation of cytosol fatty acid binding proteins. Factors which increased oleic acid's protonated concentration or diminished the unstirred layer resistance enhanced its absorption rate. On the contrary, factors which enhanced its micellar solubility or interfere with its transfer out of the cell membrane decreased its overall rate of absorption.

Human studies

Pinto Correia et al. (1963) studied the absorption of oleic acid in 15 healthy human volunteers (no further details given). The test subjects received a test meal containing 30 μCi ¹³⁹I‐C₁‐labelled oleic acid suspended in an emulsion of 90 mL milk and 10 mL olive oil, followed by 150 mL milk. Blood samples were collected 1, 2, 4, 6, 12 and 24 h after ingestion and stools for up to 120 h. Radioactivity in blood samples reached a maximal mean value of 2.8% of applied activity per litre of blood 4 h after treatment. Data on radioactivity excreted via faeces indicated that 3.1% (range: 0.03–8.6%) of applied radioactivity was detected in the faeces.

Overall, the Panel considered that caprylic‐, capric‐, oleic‐, lauric‐, palmitic‐, myristic‐ or stearic acid like other fatty acids were readily and extensively absorbed from the gastrointestinal tract. After absorption, fatty acids were either metabolised or incorporated into chylomicrons, which enter the systemic circulation. Ultimately, fatty acids, either incorporated into glycerides and phospholipids, were catabolised via the β‐oxidation pathway and the tricarboxylic acid cycle to carbon dioxide which is finally excreted via exhalation.

Caprylic acid

In the study by Litton Bionetics (1976), caprylic acid (purity 98%) was evaluated for its mutagenicity in the bacterial reverse mutation assay using Salmonella Typhimurium strains TA1535, TA1537 and TA1538, and for induction of mitotic recombination in Saccharomyces cerevisiae strain D4 using both the plate incorporation and pre‐incubation methods in the absence and presence of liver S9 metabolic activation from rat mouse and monkey. Concentrations used, selected from survival tests, were 2.5, 1.25 or 0.625 μg/plate for the bacterial reverse mutation assay and 13, 6.5 or 3.25 μg/plate for induction of mitotic recombination. Negative results were obtained for both mutagenicity and mitotic recombination. The Panel noted that the set of S. Typhimurium tester strains was limited and that the mitotic recombination assay does not belong to the genotoxicity tests recommended for regulatory purposes (EFSA Scientific Committee, 2011).

In the study by Zeiger et al. (1988), caprylic acid (purity 99%) was assessed for its mutagenicity in the reverse mutation assay using S. Typhimurium strains TA1535, TA1537, TA97, TA98 and TA100 up to a maximum concentration of 3,333 μg/plate in dimethylsulfoxide (DMSO), both in the absence and presence of rat and Chinese hamster S9 metabolic activation at 10% and 30%, and no mutagenicity was observed. The Panel noted that the study complies with current OECD Guideline No 471 (OECD, 1997) with the exception that tester strains S. Typhimurium TA102 or Escherichia coli WP2uvrA bearing AT mutation were not used.

In the study by Heck et al. (1989), caprylic acid (unknown purity) was tested for induction of gene mutation in the bacterial reverse mutation assay S. Typhimurium strains TA98, TA100, TA1535, TA1537 and TA1538 with and without S9 metabolic activation at single concentration of 5,000 μg/plate and for the induction of unscheduled DNA synthesis (UDS) in hepatocytes freshly obtained from male Fisher rats at final concentration of 300 μg/mL. Negative results were obtained in both assays. However, the Panel noted that the reporting of data was marginal and that tester strains S. Typhimurium TA102 or E. coli WP2uvrA bearing AT mutation were not used.

Capric acid

Szybalski (1958) reported that capric acid (unknown purity) in a bacterial mutation test (no details reported) did not induce streptomycin‐independent revertants of E. coli using the paper disc method. The Panel noted that the method employed does not belong to the genotoxicity tests recommended for regulatory purposes (EFSA Scientific Committee, 2011).

Oda et al. (1978) assessed the DNA‐modifying effects of capric acid at concentration of 18 μg/disk by the Rec‐assay (DNA repair test) using Bacillus subtilis mutant strain M45 rec‐, unable to repair DNA damage and the wild‐type strain H17 rec+ as control. Negative results were obtained. However, the Panel noted that the results obtained are of limited relevance since this assay does not belong to the genotoxicity tests recommended for regulatory purposes (EFSA Scientific Committee, 2011).

In the study by Zeiger et al. (1988), capric acid (purity 99%) was assessed for its mutagenicity in the reverse mutation assay using S. Typhimurium strains TA1535, TA1537, TA97, TA98 and TA100 up to a maximum concentration of 666 μg/plate in DMSO, both in the absence and presence of rat and Chinese hamster S9 metabolic activation at 10% or 30%, and no mutagenicity was observed. The Panel noted that the study complies with current OECD Guideline No 471 (OECD, 1997) with the exception that tester strains S. Typhimurium TA102 or E. coli WP2uvrA bearing AT mutation were not used.

Lauric acid

Parry et al. (1981) reported that lauric acid (no details on purity provided), in Saccharomyces cerevisiae D6, induced aneuploidy at concentrations from 10 to 200 μg/mL. However, analysis of crossing over of chromosomes in the same cell cultures revealed no effects. No data were available on cytotoxicity but it was stated that increased yeast cell lethality might influence the results on aneuploidy. The Panel noted that the test does not belong to the genotoxicity tests recommended for regulatory purposes (EFSA Scientific Committee, 2011).

In the study by Zeiger et al. (1988), lauric acid (purity 99%) was assessed for its mutagenicity in the reverse mutation assay using S. Typhimurium strains TA1535, TA1537, TA97, TA98 and TA100 up to a maximum concentration of 666 μg/plate in DMSO, both in the absence and presence of rat and Chinese hamster S9 metabolic activation at 10% or 30%, and no mutagenicity was observed. The Panel noted that the study complies with current OECD Guideline 471 No 471 (OECD, 1997) with the exception that tester strains S. Typhimurium TA102 or E. coli WP2uvrA bearing AT mutation were not used.

Myristic acid

In the study by Zeiger et al. (1988), myristic acid (purity 98%) was assessed for its mutagenicity in the reverse mutation assay using S. Typhimurium strains TA1535, TA1537, TA97, TA98 and TA100 up to a maximum concentration of 3,333 μg/plate in DMSO, both in the absence and presence of rat and Chinese hamster S9 metabolic activation at 10% or 30%, and no mutagenicity was observed. The Panel noted that the study complies with current OECD Guideline No 471 (OECD, 1997) with the exception that tester strains S. Typhimurium TA102 or E. coli WP2uvrA bearing AT mutation were not used.

Heck et al. (1989) tested myristic acid (unknown purity) in the bacterial reverse mutation assay in the presence and absence of metabolic activation using S. Typhimurium strains TA98, TA100, TA1535, TA1537 and TA1538 at concentrations up to 10,000 μg/plate. Negative results were obtained. Furthermore, myristic acid did not induce gene mutations in the mouse lymphoma assay (L5178Y TK+/− cells) at concentrations of up to 62 μg/mL (without metabolic activation) or 125 μg/mL (with metabolic activation) and UDS in hepatocytes freshly obtained from male Fisher rats at final concentration of 300 μg/mL. However, the Panel noted that the results obtained are of limited validity since only one concentration was used in both assays, reporting of data was marginal and that tester strains S. Typhimurium TA102 or E. coli WP2uvrA bearing AT mutation were not used.

Stearic acid

Parry et al. (1981) reported that stearic acid (no details provided) did not induce aneuploidy in S. cerevisiae D6 at concentrations up to 500 μg/mL. Additionally, analysis of crossing over of chromosomes in the same treated cell cultures revealed no effects. The Panel noted this assay does not belong to the genotoxicity tests recommended for regulatory purposes (EFSA Scientific Committee, 2011).

Blevins and Taylor (1982) reported that stearic acid (unknown purity) did not induce mutagenic effects in a bacterial reverse mutation assay with S. Typhimurium strains TA98, TA100, TA1535, TA1537 and TA1538 with and without rat liver metabolic activation, at a concentration of 50 μg/plate. However, the Panel noted that the results obtained are of limited validity since only one concentration was used, reporting of data was marginal and that tester strains S. Typhimurium TA102 or E. coli WP2uvrA bearing AT mutation were not used.

Crebelli et al. (1985) tested stearic acid (unknown purity) in the plate incorporation assay with S. Typhimurium strains TA98, TA100, TA1535 and TA1537 in the presence of metabolic activation (S9 from Aroclor induced male Sprague‐Dawley rats) at concentrations of 40‐1,000 μg/plate. DMSO was used as solvent. The test included positive controls. No mutagenic activity was found. The Panel noted that only treatments in the presence of S9 metabolic activation were used and that tester strains S. Typhimurium TA102 or E. coli WP2uvrA bearing AT mutation were not used.

In the study by Shimuzu et al. (1985), stearic acid (60% purity) was assessed for its mutagenicity in a bacterial reverse mutation assay using S. Typhimurium strains TA98, TA100, TA1535, TA1537, TA1538 and E. coli WP2uvrA. The pre‐incubation test was performed in the presence and absence of metabolic activation (liver S9 from polychlorinated biphenyl induced male rats) at concentrations of 1–1,000 μg/plate (seven concentrations) including negative, vehicle and positive control. The test substance did not induce an increase in revertants colonies compared to the concurrent negative control. The Panel noted that the study complies with current OECD Guideline 471 No 471 (OECD, 1997).

Oleic acid

Cotruvo et al. (1977) assessed the mutagenic potential of oleic acid (no details reported) in a bacterial reverse mutation assay in S. Typhimurium strains TA98, TA100, TA1535, TA1536, TA1537 and TA1538, and in the Saccharomyces cerevisiae D3 mitotic recombination assay with and without metabolic activation. No genotoxic activity was reported in both assays at maximum solubility of 99 μg/mL. The Panel noted that the tester strains S. Typhimurium TA102 or E. coli WP2uvrA bearing AT mutation were not used and that the mitotic recombination assay does not belong to the genotoxicity tests recommended for regulatory purposes (EFSA Scientific Committee, 2011).

Mortelmans et al. (1986) tested oleic acid (technical grade) in the bacterial reverse mutation assay, in S. Typhimurium strains TA98, TA100, TA1535 and TA1537 in the presence and absence of metabolic activation (S9‐mix from hamster or rat liver). The test substance was dissolved in DMSO and tested up to the toxicity threshold (3.3–10 μg/plate without metabolic activation and 333 μg/plate with metabolic activation). No mutagenic activity was observed. The Panel noted that the tester strains S. Typhimurium TA102 or E. coli WP2uvrA bearing AT mutation were not employed.

Parry et al. (1981) reported that oleic acid (no details provided) induced aneuploidy in S. cerevisiae D6 at concentrations from 100 to 500 μg/mL. However, analysis of crossing over of chromosomes in the same cell cultures revealed no effects. No data were available on cytotoxicity but increased yeast cell lethality might influence the results on aneuploidy. The Panel noted that the test does not belong to the genotoxicity tests recommended for regulatory purposes (EFSA Scientific Committee, 2011).

Osawa and Namiki (1982) assessed the DNA‐modifying effects of oleic acid (unknown purity) at concentration of 100 or 1,000 μg/disk by the Rec‐assay (DNA repair test) using Bacillus subtilis mutant strain M45 rec‐, unable to repair DNA damage and the wild‐type strain H17 rec+ as control. Negative results were obtained. However, the Panel noted that the results obtained were of limited relevance since this assay does not belong to the genotoxicity tests recommended for regulatory purposes (EFSA Scientific Committee, 2011).

Kinsella (1982) studied the induction in vitro of forward mutations after exposure to oleic acid by the mammalian cell gene mutation assay in V79 6‐thioguanine resistant Chinese hamster lung fibroblasts without metabolic activation. At a concentration of 1 μg/mL, no increase in the mutation frequency compared to the negative concurrent control was observed. In further experiments using also V79 cells, sister chromatid exchange (SCE) analysis and quantitation of induced chromosomal aberrations in growing cells exposed simultaneously to 5‐bromo‐2’‐deoxyuridine and the test agents were performed. No DNA damage and no clastogenic effects were detected at concentrations of 2.5, 5 or 10 μg oleic acid/mL. However, the Panel noted that the study design had significant limitations which included single concentration in the mammalian cell gene mutation assay, very low concentrations (1–10 μg/mL), the absence of treatment in the presence of S9 metabolic activation and no indication of sampling time. On this basis the results obtained were not considered reliable for risk assessment.

Shimuzu et al. (1985), in a bacterial reverse mutation assay using S. Typhimurium strains TA98, TA100, TA1535, TA1537 and TA1538, and E. coli WP2uvrA assessed the mutagenicity of oleic acid (99.9% purity) dissolved in acetone. The pre‐incubation test was performed in the presence of metabolic activation (liver S9 from polychlorinated biphenyl induced male rats) at concentrations of 1‐5,000 μg/plate (eight concentrations) including negative, vehicle and positive control. In TA1537 and TA1538, the test substance was tested up to the cytotoxic doses (100 μg/plate). The test substance did not induce any increase in revertant colonies compared to the concurrent negative control. The Panel noted that the study complied with current OECD Guideline No 471 (OECD, 1997).

In the study by Higgins et al. (1999), oleic acid (99% purity) was assessed for the induction of the SCE in the Indian muntjac fibroblasts, complexed with 2% bovine serum albumin without metabolic activation at a single concentration of 50 μM. Analytical control revealed significant uptake of oleic acid by the cells but no genotoxic effects were reported. The Panel noted severe limitations of the study protocol which included the use of a single concentration, treatments performed only in the absence of S9 metabolism, inadequate selection of the highest concentration which was low and did not induce appropriate reduction of mitotic index. In addition, the SCE assay does not belong to the genotoxicity tests recommended for regulatory purposes (EFSA Scientific Committee, 2011).

Rats

Young rats (n = 4/sex per group) were fed 16 weeks premating 15% oleic acid in the diet (equivalent to 7,500 mg/kg bw per day). The animals were mated within the test group and four litters were born (Carroll and Noble, 1957). The study was very limited (number of animals and reporting). The Panel considered that this study cannot be used for the hazard characterisation.

Rats

Narotsky et al. (1994) studied the developmental toxicity of caprylic acid in a screening test (a limited number of parameters were tested). Pregnant Sprague–Dawley rats were administered by gavage with caprylic acid (purity 99.5%) in corn oil on gestation days (GD) 6–15 at dose levels of 0 (20 animals), 1,125 mg/kg bw per day (16 animals) or 1,500 mg/kg bw per day (16 animals). The dams were allowed to deliver and their litters were examined through post‐natal day 6. The treatment induced maternal toxicity: at the low dose level, rales were observed in 15/16 rats and dyspnoea in 4/16; similar incidences were found at the high dose. Five low‐dosed rats and seven high‐dosed rats died during the treatment period while no deaths were observed in the controls. Maternal body weight gain was significantly and dose dependently reduced. According to the authors the maternal effects might be influenced by application error; the authors stated that ‘death may have been due to paragastric intubation’. The treatment did not affect the number of implantations/dam, perinatal loss, number of live pups at post‐natal day 1 and pup weight at post‐natal day 1 and 6. At the high dose level (1,500 mg/kg bw per day) but not at 1,125 mg/kg bw per day, the number of live pups at post‐natal day 6 was statistically significantly reduced. This decreased viability occurred only in litters of dams with peripartum respiratory symptoms. External malformations were not observed. Overall, the Panel noted that in this study no developmental toxicity occurred at a maternal toxic dose of 1,125 mg/kg bw per day. Developmental toxicity at a dose of 1,500 mg/kg bw per day was related to maternal toxicity which may have been induced by application error rather than by the test item. However, the Panel noted that the oral limit dose according to OECD Guideline 414 is 1,000 mg/kg bw per day. The Panel also considered that the study design of this study was too limited to conclude on prenatal developmental toxicity.

In the study of Scott et al. (1994), a group of 12 pregnant Sprague–Dawley rats received via gavage a single dose of 2,700 mg caprylic acid (undiluted)/kg bw at GD 12. The control group (10 animals) was untreated. Caesarean section was performed at GD 20 and implantations, resorptions and dead fetuses were counted. The fetuses were removed, sexed, weighed and examined for external malformations. Two‐third of the fetuses were examined for internal malformations and the others for skeletal malformations. The treatment resulted in severe maternal toxicity (no further information provided) but did not induce any developmental toxicity except a decrease in fetal weight (no data about statistical significance given) which was attributed to maternal toxicity. Based on these findings, the authors concluded that caprylic acid did not induce developmental toxicity. The Panel noted that the design of this study was very limited (low number of animals and inappropriate dosing regimen).

The Panel noted that insufficient data were available on reproductive toxicity of fatty acids. Limited data on developmental toxicity were available only for caprylic acid.

Overall, the Panel considered that data on reproductive and developmental effects were too limited to reach a conclusion on these endpoints.