Safety and efficacy of saponified paprika extract, containing capsanthin as main carotenoid source, for poultry for fattening and laying (except turkeys) (2024)

EFSA J. 2020 Feb; 18(2): e06023.

PMID: 32874232

3.1.1.1. Manufacturing

The saponified paprika extract is manufactured by a two‐step procedure involving first the extraction of the dried fruits from C. annuum L. (sweet peppers or paprika) with hexane, and second the saponification using potassium or sodium hydroxide, ■■■■■ As a result of the saponification, the carotenoid esters, triglycerides and other lipid‐soluble esters present in the paprika extract are hydrolysed into free carotenoids, fatty acids and glycerol. Antioxidants are added several times during manufacturing to protect carotenoids against oxidation.

3.1.1.2. Composition

Data were provided from three companies.¹³ The analysis of a total of seven batches of saponified paprika extract showed mean values of saponification rate of 82.6% (Company A, two batches), 78.4% (Company C, two batches) and 77.5% (Company E, three batches).¹² The concentration of TC was determined in four batches of each company and ranged between 25.0 and 55.5 g/kg of which 36.9–57.8% was capsanthin. Additional three batches from company A were analysed and showed higher values of TC (75.2–76.8 g/kg; capsanthin 41.8–44.7% of TC). These higher values are due to the variability of the raw material that is affected by growth conditions of the plant (C. annuum) including yearly changes due to natural variation in precipitation and soil conditions or occasionally changes in cultivation practices. The analysed batches were all within the proposed specifications (TC: 25–90g/kg; capsanthin: >35% of TC).¹²

The carotenoid profile was analysed in one batch of two companies (Table1).¹² The major carotenoids in these two saponified extracts, further to capsanthin, were zeaxanthin, β‐carotene and cucurbitaxanthin A. Capsorubin was present in small quantities.

The applicant provided results of analysis of zeaxanthin, lutein and β‐carotene in further batches of the saponified extract (Table2).

Information on the content of protein, ash, water (Table3) and fatty acids (Table4) were provided in four batches per company.¹⁴

Fatty acid distribution was determined in two batches of company A and C and one batch of company E. Major fatty acids in the extracts are linoleic acid (C18:2), palmitic acid (C16:0), oleic acid (C18:1), linolenic acid (C18:3), myristic acid (C14:0) and stearic acid (C18:0); they account for at least 92% of total fatty acids (Table4). The fat content of the extract was not provided.

■■■■■^{■■■■■}■■■■■^{■■■■■}■■■■■

The mineral fraction was characterised by high amounts of potassium (up to 93.7 g/kg) due to the use of potassium hydroxide in the saponification.¹⁷

3.1.1.3. Purity

During the manufacturing process, hexane is used in the extraction. The content of hexane was analysed in at least three batches per company. Results were below the International Cooperation on Harmonization of Technical Requirements for Registration of Veterinary Medicinal Products (VICH) Guideline limit (290mg/kg, Class 2 solvent). The highest value was 130mg/kg in one batch; other values were <25mg/kg.¹⁷

Data on residues of other solvents were submitted. Considering that the extract will be marketed in more diluted forms, the data did not raise any concern. This was not the case for one batch from one company in which the concentration of benzene was 64mg/kg which exceeds the VICH limit for this Class 1 solvent (2mg/kg) and according to VICH guidelines, substances of Class 1 should be avoided in active substances, excipients and preparations. The company involved in the production confirmed that benzene is not used in any step of the manufacturing and HACCP procedure is in place to control solvent levels. In addition, analytical data for additional batches were submitted that showed values of benzene <1mg/kg.¹⁸

Capsaicin and closely related compounds summarised as capsaicinoids, the pungent principles in C.annuum, were present at levels <300mg/kg in the saponified paprika extracts.¹⁹ The applicant provided analysis of three batches of a feed additive preparation (Company A) to show that capsaicin were either not detected (in two batches; limit of detection (LOD): 0.1mg/kg) or was present at low concentration (13.5mg/kg in one batch).²⁰ The characteristics of capsaicin were described in the EFSA opinion on the use of colouring agents in animal nutrition (EFSA, 2006) and in EFSA opinion on the re‐evaluation of paprika extract as food additive (EFSA ANS Panel, 2015). The FEEDAP Panelnotes that according to Regulation (EC) No231/2012, concentration of capsaicin is limited to ‘not more than 250 mg/kg’ in the food additive (paprika extract, E 160c). The same specification is recommended for the feed additive. Considering that the extract is formulated (and diluted) to obtain the final additive, the values reported by the applicant are of no concern.

Heavy metals and arsenic were measured in five different batches from three companies. Three batches showed values of Pb <1.0mg/kg, Cd <0.1mg/kg, Hg <0.01mg/kg and As <1.0mg/kg. Values of two other batches were Pb <0.05mg/kg, Cd <0.01mg/kg, Hg <0.005mg/kg and As <0.1mg/kg.¹⁷

Aflatoxins (B1, B2, G1 and G2) were measured in one batch of each company. All values were <0.1μg/kg in two batches. In the third batch, aflatoxins were not detected (LODs were <7.5, 2, 2 and 5μg/kg).¹⁷

The applicant stated that dioxins (polychlorinated dibenzodioxins (PCDDs) and polychlorinated dibenzofurans (PCDFs)), dioxin‐like polychlorinated biphenyls (PCBs) and non dioxin‐like PCBs are regularly monitored in the starting material (unsaponified oleoresin).²¹ Data on dioxins in saponified oleoresin were provided from two companies: sum of dioxins were 2.27ng (Comp A) and 8.09ng WHO‐PCDD/F‐TEQ/kg (Comp E).²²

Pesticides (>300 substances) were analysed in one batch of saponified paprika extract from one company and in the non‐saponified extract from another company.²³ Traces of buprofezin, etofenprox, diafenthiuron, diphenylamine or biphenyl were detected. However, considering that the additive is not marketed as such but in form of diluted, standardised and stabilised feed additive preparations, such amounts found in the saponified extract are not expected to pose a risk to target animal or consumer health. No pesticide residues were detected in a commercial powder preparation of the same company.

3.1.1.4. Physicochemical properties of capsanthin

Capsanthin ((3R,3’S,5’R)‐3,3’‐dihydroxybeta,kappa‐caroten‐6’‐one, CAS number 465‐42‐9) has the molecular formula C₄₀H₅₆O₃ and the molecular weight of 584.87 g/mol. Its structural formula is given in Figure1.

Physicochemical data of capsanthin were provided by the applicant and was also described in the EFSA opinion on the use of colouring agents in animal nutrition (EFSA, 2006).

3.1.2.1. Typical composition

The saponified extract is not used directly as a feed additive. It is diluted, standardised and stabilised before placing in the market. The applicant reported that such products typically contain 0.5–5.0% TC and 0.2–2.0% capsanthin. The preparations may be solid or liquid and may contain carriers (e.g. calcium carbonate, Tagetes meal bagasse, rice hulls, sesame meal, wheat bran, corncob meal, coconut shells, water), anticaking agents (e.g. silicic acid, colloidal silica and kaolinitic clay) and antioxidants.

3.1.2.2. Physical state

Particle size data obtained by sieve analysis were submitted for five preparations from the three companies. For company A, data showed that the amount of particles <50μm ranged between 45 and 55% (two preparations a total of six batches), no particles <10μm were found. Company C (three batches) reported that 5–11% of the particles were smaller than 63μm and 3–6% were bigger than 495 μm. The third company reported that 98% of the particles were <850μm.

The dusting potential was determined in four different preparations (at least one per company) using the Stauber–Heubach method.²⁴ Results were <0.1 (Company C), 0.3 (Company E), 20.9 (Company A) and 21.4/58.2 g/m³ (Company A, two batches).

3.1.2.3. Stability and hom*ogeneity

The sensitivity of carotenoids to temperature, light and oxygen requires the inclusion of antioxidants at the different steps of manufacturing of the active substance and the preparations of the additive to assure the stability of capsanthin.

Metabolic studies

The metabolism of capsanthin was assessed by the FEEDAP Panelin 2006 (EFSA, 2006). No new data have been supplied by the applicant for the current assessment and, to the knowledge of the FEEDAP Panel, no specific data on capsanthin have been published since 2006. A summary of the information available on the metabolism of capsanthin in the target species, laboratory animals and humans is given as follows: (i) no data are available concerning the metabolic fate of capsanthin in poultry or laboratory animals; only evidence is given that the bioavailability of capsanthin from esterified capsanthin and from saponified (free) capsanthin from paprika extract in the chicken, measured through the plasma levels of free capsanthin, are similar, implying ester hydrolysis prior to absorption, (ii) capsanthin is eliminated in its free form in the egg yolk; the metabolic fate in laying hen of synthetic (3S,5R,3'S,5'R)‐capsorubin, a chemically close compound harbouring two cyclopentanyl end rings instead of one in capsanthin, identified 14 metabolites with almost identical chromophoric values but diverse polarities ranging from monosulfate to fatty acid monoesters; it can be assumed that a similar metabolic behaviour of the common structural part of capsanthin would occur, (iii) in humans, the bioavailability of capsanthin from paprika extract was shown to be very low; after administration of capsanthin from paprika juice for one week, capsanthin was distributed in plasma lipoproteins and its concentration plateaued between day 2 and day 7 to become undetectable after day 16. Humans may have the potential metabolic activity for the oxidation of secondary hydroxyl groups in various xanthophylls, i.e. the formation of capsanthon from capsanthin by the oxidation of the 3′‐hydroxyl group to the 3′‐keto group (Etoh etal., 2000).

In addition, it is noted that absorption and bioavailability of carotenoids will strongly depend on the matrix in which they are administered and thus may differ between unsaponified and saponified preparations and will be influenced by the presence of solubilisers in the additive.

3.2.1.1. Deposition of capsanthin in chicken tissues

Limited data reported in the former EFSA assessment (2006) on the deposition of capsanthin in chicken tissues refer to a study of Erkek etal. (1993) where chickens administered a diet supplemented with 20mg xantophylls/kg exhibited capsanthin residues in the skin amounting to 0.4mg/kg.

New residue data (total carotenoid and capsanthin residues) were measured in tissues of chickens for fattening from the tolerance study (see Section 3.2.3.1).³² Four groups of 1‐day‐old chickens (20 birds/replicate, 6 replicates/treatment) were administered for 36days a feed supplemented with 0, 40, 80 and 240mg total carotenoids/kg in the form of a saponified paprika extract. The analysed values (given as mean of starter and grower diets at start) were 2.7, 40.0, 77.2 and 250.0mg TC/kg (0.06, 17.5, 35.0 and 104mg capsanthin/kg). Tissues, except those from the highest dose group, were sampled at slaughter (day 36) and kept frozen until analysis. Total carotenoids and capsanthin were determined by high‐performance liquid chromatography (HPLC), with an LOD of 0.5 mg carotenoid/kg wet tissue. The results are shown in Table5.

3.2.1.2. Deposition of capsanthin in the egg yolk

The applicant submitted three studies assessing the deposition of dietary carotenoids from paprika extracts in the egg (Hamilton etal., 1990; Lai etal., 1996; González etal., 1999). These papers were already assessed by the FEEDAP Panel(EFSA, 2006) and considered insufficient to be used for the assessment of consumer exposure because of different analytical methods used, different background contents of diets, lack of additive identity and weaknesses in reporting.

New residue data (total carotenoid and capsanthin residues) were measured in whole egg samples from the tolerance study in laying hens (see Section 3.2.3.2).³³ Four groups (seven hens per replicate, eight replicates) of 24‐week‐old laying hens were administered for 56days a feed supplemented with 0, 15, 80 and 240mg total carotenoids/kg in the form of a saponified paprika extract. The analysed values were 4.8, 20.5, 83.4 and 265mg TC/kg (<1, 8.2, 40.4 and 128 mg capsanthin/kg). All eggs, except those from the highest dose group, laid on day 55 were collected per replicate. Liquid eggs were pooled per replicate and stored frozen until analysis. Total carotenoids and capsanthin were determined by HPLC, with an LOD of 0.5mg carotenoid/kg whole egg. The results are shown in Table6.

3.2.2.1. Genotoxicity studies including mutagenicity

Saponified paprika extract was tested in a bacterial reverse mutation assay on Salmonella Typhimurium strains TA1535, TA1537, TA100 and TA98 and on Escherichia coli strain WP2uvrA both in the absence and presence of S9‐mix (rat liver S9‐mix induced Aroclor 1254) in compliance with OECD test guideline 471.³⁴ The vehicle of the test item was ethanol and both the plate incorporation and the pre‐incubation methods were applied. Precipitation was reported at 5000 μg/plate and sporadically at 1600 μg/plate. In the S. Typhimurium tester strains, the maximum analysable concentration was in the range 17–52 μg/plate or mL and was limited by cytotoxicity. In the Escherichia coli tester strain WP2uvrA, no toxicity was observed up to and including the dose level of 512 μg/mL. No increase in the number of revertant colonies was reported in any experimental condition while the positive controls performed as expected.

The possible clastogenicity and aneugenicity of saponified paprika extract was tested in an invitro micronucleus assay in cultured peripheral human lymphocytes, in compliance with OECD test guideline 487.³⁵ Two independent experiments were performed with and without metabolic activation (phenobarbital and ß‐naphthoflavone induced rat liver S9‐mix). Ethanol was used as the vehicle. In the first cytogenetic assay, the test item was tested up to 300μg/ml for a 3‐hour exposure time with a 27‐hour harvest time in the absence and presence of S9‐fraction. At this concentration, 13 and 31% cytostasis was reported with and without metabolic activation, respectively, and the test substance formed a precipitate in the culture medium.

In the second cytogenetic assay, the test item was tested up to 300μg/mL for a 24‐hour exposure time with a 24‐hour harvest time in the absence of S9‐mix. The top concentration caused 56% cytostasis and the formation of precipitate in the culture medium. No statistically significant or biologically relevant increase in the number of mono‐ and binucleated cells with micronuclei was reported in the absence and presence of S9‐mix, in either of the two experiments. The positive control chemicals produced a statistically significant increase in the number of binucleated cells with micronuclei.

3.2.2.2. Subchronic toxicity

Groups of 10 male and 10 female Wistar rats were fed diets containing 0 (control), 1,000 (low dose), 5,000 (mid dose) or 10 000 (top dose) mg test saponified paprika extract³⁷ /kg feed for 90days in a toxicological study in compliance with OECD test guideline 408.³⁶ The groups of male rats received mean dosages of the test item of 0, 84, 428 and 858 mg/kg body weight (bw) per day over the course of the study; the corresponding dosages for females were 0, 85, 448 and 879mg/kg bw per day. Clinical observations were made daily on all animals with more detailed ‘arena observations’ being made weekly. Body weights and food consumption were measured weekly. Functional tests for pupillary reflex, hearing ability, static righting reflex, grip strength and locomotor activity were performed on 5 animals/sex per group towards the end of the study. Ophthalmic observations were made on all animals at the beginning and end of the study period. Daily vagin*l lavage samples from all females were taken towards the end of the study and were examined for cytology for oestrus cycle determination. Blood samples were taken from all rats on the day of necropsy and were used for haematological and blood biochemical examination. All animals were subjected to necropsy, organs were weighed and tissues preserved for histological examination. Tissues and organs from all animals of the control and high‐dose groups were independently examined microscopically by two pathologists, with particularly detailed examination of the testes.

No mortality occurred. No treatment‐related clinical signs were seen other than red colouration of the faeces of mid and top dose group rats. Food consumption and body weights were not affected by the treatments. The only effect on functional observations was a significantly lower forelimb grip strength in mid and top dose males as compared with male controls. However, there was no effect on grip strength of the forelimbs in females and no effect on hindlimb grip strength in either sex. There were no treatment‐related effects on ophthalmology results or oestrus cycle length. Haematology revealed no effects in females. In males, there were small but significant differences from control values in some haematological parameters (white blood cell and lymphocyte count decreased at the top dose, and mean corpuscular haemoglobin concentration decreased at all doses) in some treated groups, but in the absence of a dose–response relationship these were not regarded as treatment‐related. There were also some small but significant differences from control values in some blood biochemistry parameters (elevated sodium in males at all dose levels and in mid‐dose females, decreased glucose in low‐dose females and elevated cholesterol in mid‐ and high‐dose females). The only blood biochemistry changes that showed a dose–response relationship were those of cholesterol in females. The absence of corroborative findings in the opposite sex and/or the absence of a dose–response relationship suggested that none of the statistically significant changes in blood biochemistry parameters were caused by the treatment. The only treatment‐related effect seen at necropsy was a yellowish discolouration of the adipose tissue in some males in the mid‐dose group and all males and females in the top dose groups. There were no treatment‐related effects on organ weights or on histopathology.

The FEEDAP Panelconcludes that no adverse effects were caused in rats given the saponified paprika extract at dietary concentrations of up to the highest dose tested of 10 000 mg/kg (up to 858 mg (41.4 mg TC)/kg bw per day in males up to 879 mg (42.4 mg TC)/kg bw per day in females).

3.2.2.3. Conclusions

The saponified paprika extract assessed is not genotoxic in a bacterial reverse mutation assay and in an invitro mammalian cell micronucleus test. No adverse effects were caused in rats given the paprika saponified extract under assessment at dietary concentrations up to the highest dose tested of 10 000mg/kg (up to 858 mg (41.4mg TC)/kg bw per day in males up to 879mg (42.4mg TC)/kg bw per day in females). The NOEL from the study is 41.4mg TC/kg bw per day for saponified paprika extract. The FEEDAP Panelis aware that the above results are obtained with one saponified paprika extract, however, they could be considered representative of other saponified paprika extracts with a TC content in the range of 25–90g/kg and a capsanthin content of >35% of TC. The FEEDAP Panelalso notes that reproduction toxicity studies were not provided.

3.2.3.1. Tolerance study in chickens for fattening

A total of 480 1‐day‐old male chickens for fattening (Ross 308) was distributed in 24 pens of 20 birds and allocated to four treatment groups (representing six replicates per treatment).³² The test item was a saponified paprika extract on a mineral carrier containing 21.35g TC/kg and 8.5g capsanthin/kg (expressed as 40% TC). The groups were fed diets supplemented with 0 (control), 40 (1× maximum recommended level), 80 (2×) or 240 (6×) mg TC/kg complete feed for 36days. The pelleted corn–soybean‐based diets (methionine hydroxy analogue supplemented) were given as starter (day 1–21; 18.9% crude protein (CP), 12.8 MJ metabolisable energy (ME)/kg, 0.94% methionine‐cysteine (met‐cys) and grower diet (day 22 until study completion; 18.0% CP, 13.2 MJ ME/kg, 0.86% met‐cys). Total carotenoids and capsanthin were analysed in the feed before and at the end of the two feeding periods, the results are given in Table7.

Health status, including mortality, was monitored daily, body weight and feed intake (per replicates) was measured on day 22 and 36 and feed to gain ratio was calculated for the corresponding periods. Blood samples for haematology³⁸ and clinical chemistry³⁹ were taken at the end of the study from two randomly selected birds/pen (12 per treatment). The same birds were killed for necropsy, organ weight determination (liver, spleen, kidneys, heart) and histological examination of liver, kidney, muscle and skin/fat samples. Another two birds per replicate were killed for edible tissue sampling (liver, kidneys, breast muscle and skin+fat (from breast area)).

Statistical evaluation was made by analysis of variance (ANOVA) of data from a randomised complete block design (with blocks (four adjacent pens) as random effect). The experimental unit was the pen. Group means were compared with Tukey test (adjusted for continuous variables). Mortality data were assessed using categorical data analysis (Wilcoxon rank sum, Kruskal–Wallis test). A significance level α=0.05 was used.

Mortality was low (average 1.9%) and not treatment related. No differences between the groups were detected by statistical analysis for the final body weight (mean: 2,465 g), daily feed intake (mean: 98 g) and feed to gain ratio (mean 1.46).

No significant differences were detected among treatments in serum enzyme activities, protein and lipid related variables, or electrolytes except for bilirubin, creatinine and alanine aminotransferase (ALT). Bilirubin level was 0.115mg/dL for the control group, whereas it was below the LOD of 0.05mg/dL for all groups treated with paprika extract. Since no adverse consequences of lower bilirubin levels are known, this effect (if present) appears to be without biological relevance for the safety of target animals. Creatinine significantly decreased with the inclusion of saponified paprika extract (from 0.214mg/dL to 0.180, 0.158 and 0.141 for the groups with 40 (1×), 80 (2×) and 240 (6×) mg TC/kg supplementation level of the test item); these changes are not considered to be of any safety relevance for poultry. ALT was only lower in the intermediate level group (1.7 U/L) compared to the control group (2.5 U/L). No significant differences were seen between the groups for haematological variables, cell counts or differential cell counts.

No significant differences were found in organs weights of birds expressed in relative weight.

The high‐dose group produced yellowish to orange coloured fat, skin, subcutaneous tissue and liver, but this was not associated with any gross pathology or histopathology.

3.2.3.2. Tolerance study in laying hens

The tolerance study in laying hens was performed using a saponified paprika extract on a mineral carrier as a test item containing 20.8mg TC/g and 7.7 mg capsanthin/g (expressed as 37% TC). A total of 224 laying hens (Hy‐Line Brown, 24weeks old, 1.83 kg bw) was allocated to four treatment groups fed diets supplemented with 0 (control), 15, 80 (2×) or 240 (6×) mg TC/kg complete feed for 56 days according to a randomised complete block design.³³ Group size was eight replicates (enriched laying hen cages) with seven hens each. Blocking factors were 2‐wk previous laying rate and battery side, initial body weight was also considered when building groups.

The diet, given in mash form, consisted mainly of maize and soybean meal and was supplemented with MHA and tryptophan. The basal diet was calculated to provide 14.9% CP, 0.58% met+cys and 11.1 MJ ME/kg. The analysed TC levels in the diets were initially 4.8, 20.5, 83.4 and 265mg TC/kg for the control diet and the diets with nominal 15, 80 and 240mg TC/kg feed, the capsanthin contents were <1, 8.2, 40.4 and 128mg/kg, respectively. These concentrations were measured 4 months afterwards as follows: 4.7, 21, 82 and 197mg TC/kg, and <1, 7.2, 38.6 and 98.8mg capsanthin/kg. An essential reduction of the test substance during the 2‐month experimental period could therefore only be supposed for the high‐dose group, its magnitude is probably less than 15%.

Health status including mortality was monitored daily. Animals were weighed by replicate at the beginning and end of the study. Feed consumption was calculated for both 28‐day periods. All eggs laid were counted and weighed per replicate every second day on weekly days. Performance variables (laying rate, average egg weight and mass, average daily feed intake and average feed to egg mass ratio) were calculated. At the end of the study, two birds/replicate (16 per treatment) were selected for blood sampling (haematology⁴⁰ and clinical chemistry³⁹) and killed thereafter for gross pathology examination, organ weight determination (heart, liver, spleen, kidneys) and histological examination of liver, kidney, muscle and skin/fat. Egg quality (yolk colour and Haugh Units, shell and yolk percentages, shell index and egg classification) was assessed at the end of each period. Whole egg samples were collected to measure total carotenoid and capsanthin residues.

Statistical evaluation (comparison of Tukey‐adjusted means) was made by ANOVA of data from a randomised complete block design (with blocks as random effects). Statistical significance was declared at p<0.05.

No mortality occurred during the study. No differences between the experimental groups were seen for the zootechnical endpoints: initial body weight (mean: 1,828g), final body weight (mean: 1,936g), average daily feed intake (mean: 122g), laying rate (mean: 94.9%) and egg weight (mean: 58.8g). Daily egg mass showed differences near to significance (p=0.052) and significant differences for feed to egg mass ratio (p=0.028); highest daily egg mass (56.9g) and lowest (best) feed to egg mass ratio (2.12) were found in the control group, lowest daily egg mass (54.1g) and highest feed to egg mass ratio (2.26) were found in the low TC group, for the intermediate and the high level groups both parameters were between the extremes and not significantly different from those. These differences are considered accidental.

No differences were detected for haematology and blood biochemistry with the exception of bilirubin. For bilirubin, the following blood concentrations were measured: 0.15mg/dL in the control and 0.08mg/dL⁴¹ in the group with 15mg TC/kg feed, respectively. All bilirubin values of the two groups with higher dietary TC concentrations were below LOD (0.05mg/dL); this effect appears to be without biological relevance for the safety of target animals. Creatinine (p=0.054), and mean corpuscular volume (MCV) (p=0.051) showed nearly significant differences. Creatinine in the overdose group (0.280mg/dL) was lowest and in tendency (p<0.10) different from the highest value measured for the intermediate level group (0.297mg/dL), the values for the control and the low‐level group being in between. MCV was lowest in the control groups (115.5 fL) and highest (127.8 fL) in the low‐level group, the other groups not being different from both (p<0.10).

There were no statistically significant differences detected in relative organs weights (means for heart: 0.521%, for liver: 2.040%, for spleen: 0.114% and for kidneys: 0.671%). Histopathological assessment of organs and tissues did not result in adverse findings.

3.2.3.3. Conclusions on safety for the target species

Saponified paprika extract was tolerated by chickens for fattening and laying hens at the highest level tested (240mg TC/kg feed). The FEEDAP Panelconcludes that maximum recommended use level of 40mg TC from the saponified paprika extract/kg feed is safe for chickens for fattening and laying hens. The margin of safety is at least 6. This conclusion is extrapolated to minor poultry species for fattening and laying.

3.2.4.1. Determination of a safe concentration

Based on the available data, the FEEDAP Panelconcludes that no adverse effects were caused in rats given a saponified paprika extract at dietary concentrations of up to the highest dose tested of 10,000mg/kg (up to 858mg (41.4mg TC)/kg bw per day in males up to 879mg (42.4mg TC)/kg bw per day in females) in a 90‐day study. The NOEL from the study is 41.4mg TC/kg bw per day.

3.2.4.2. Consumer exposure

The chronic exposure of consumers to residues of TC from the saponified extract of paprika in tissues of chickens end eggs of laying hens is calculated (Table9) following the methodology described in the Guidance on the safety of feed additives for consumers (EFSA FEEDAP Panel, 2017b) and using the Feed Additives Consumer Exposure (FACE) calculator available on EFSA's website (for further details see AppendixA). The residue data originating from tolerance studies (see Sections3.2.2 and 3.2.3) are summarised in Table8.

The results showed that the highest chronic exposure was for the age class ‘other children’ with 0.0562 mg/kg bw per day, the lowest for the age class ‘elderly’ with 0.0173mg/kg bw per day (for detailed results per age class, country and survey see AppendixA, TableA.1).

3.2.4.3. Safety assessment

Based on the NOEL of the 90‐day study and the exposure estimates, the Panelconsidered that there would be an adequate margin of exposure (MOE) (between 700 and 2,000) to conclude that the level of exposure to residues of the saponified paprika extract in animal tissues and products does not raise concern for the safety for consumers (EFSA Scientific Committee, 2019). The MOE is enough wide to account for the uncertainty in the inter‐ and intraspecies extrapolation and for the absence of reproduction toxicity studies.

3.2.4.4. Conclusions on safety for the consumer

The FEEDAP Panelconcludes that the use of the saponified paprika extract (TC: 25–90g/kg, capsanthin not less than 35% of TCs) for poultry for fattening and laying would not be of concern for the consumer.

Table 10

Colour of breast skin and foot pads after 42 days of feeding different levels of saponified paprika extract to chickens for fattening

Table 11

Hunter colour reflectance values of shanks and skin of chickens for fattening after 3weeks feeding of a diet supplemented with 20mg TC from paprika meal/kg

Table 12

Concentration of red pigments in plasma (μg/mL) and skin yellowness (b*) and redness (a*) values of birds fed diets containing various levels of natural pigments

Table 13

Effect of supplementing layer diets with saponified paprika extract on egg yolk colour after 56 days of feeding