Buyer protection
127 scientific publications from 15 countries, read and summarised for you – from laboratory experiments to studies in humans. Each with its original source, including findings that do not favour donkey milk.
Ongoing research project in Kiel – with our milkItaly leads with the clinical studies and the FortiLat series; China has caught up strongly with animal studies. Choose a country.
Assigned by location of the lead institution; lines show collaborations. Not a complete record of research worldwide.
Each dot is a study, coloured by study type. A click takes you to the summary.
Milk samples, proteins or cells. Shows properties and mechanisms – not what happens in the body.
Mostly mice or rats, often with artificially induced changes. Not directly transferable to humans.
Clinical studies under medical supervision. The most informative – but only for the group and product examined.
Important: This page summarises publications and describes what researchers have observed. It is not a promise of healing and not a statement about the effect of our products. Our donkey milk is a food. If you have health questions, please consult your doctor.
Constituents, processing, feeding and bacteria – including findings that do not favour donkey milk.
Laboratory digestion with human enzymes: lysozyme and α-lactalbumin remained fairly stable, and inhibition of Listeria remained measurable.
Tidona et al. · Int. Dairy J. · Original source ↗
In the laboratory, raw milk from Serbian Balkan donkeys inhibited E. coli and Salmonella, depending on temperature. 1.31 g/L lysozyme.
Šarić et al. · Int. Dairy J. · Original source ↗
94 proteins identified; around 41 % have antimicrobial or immune-related functions described in the literature, such as lysozyme.
Spada et al. · Journal of Proteomics · Original source ↗
Milk from 23 Greek jennies analysed. In the agar test, it inhibited Staphylococcus haemolyticus and Listeria monocytogenes.
Massouras et al. · Animals · Original source ↗
137 samples from 33 jennies over 21 months: no Salmonella and no Listeria were found in the study samples. 2.63 g/L lysozyme.
Šarić et al. · Animals · Original source ↗
Raw milk inoculated with bacteria and kept at 8 °C: Listeria multiplied more in donkey milk; for S. aureus there was no difference from cow’s milk.
Daminelli et al. · Veterinaria Italiana · Original source ↗
15 % condensed donkey milk applied to the teats of 48 dairy cows: clostridia fell by 55.6 %, and by 28.0 % with the comparison product.
Adduci et al. · J. Dairy Sci. · Original source ↗
Ongoing basic research in Kiel with donkey milk from DON DONKEY’S®: lysozyme, sialidases and exosomes compared with six types of milk.
Prof. Siebert · RI-B-NT Kiel · ongoing · Original source ↗
3,856 peptides in raw and 3,875 in fermented donkey milk. In the cell experiment, peptides from the fermented milk lowered the messenger substance NO.
Ning et al. · Food Res. Int. · Original source ↗
Strain from raw donkey milk: fermented milk inhibited ACE more strongly in the test tube (IC50 3.17 instead of 7.23 g/L). Blood pressure was not measured.
García-Martín et al. · Int. J. Food Microbiol. · Original source ↗
β-casein from three types of milk in the laboratory: the donkey had the smallest particles (118 nm); in the simulated intestinal phase, breakdown lay between goat and cow.
Liu et al. · Food Chem. X · Original source ↗
10 clinical isolates, 7 multi-resistant: less growth in donkey milk than in cow’s milk and nutrient broth, strongly dependent on the isolate.
Trudić et al. · Animals · Original source ↗
Raw milk from 12 jennies, 24 h at 4 °C: 14 isolates, mostly Leuconostoc. Two strains acidified well and carried no acquired resistance genes.
Bulnes et al. · Appl. Microbiol. · Original source ↗
6 samples each of donkey, mare’s, camel, human and pig milk: the fat profile of donkey milk was close to mare’s milk. We have critically reviewed the authors’ health statement.
Zhang et al. · Food Chem. X · Original source ↗
16 jennies, 8 weeks: more lucerne in the feed was associated with more alpha-linolenic acid and a lower n-6/n-3 ratio.
Liang et al. · Animals · Original source ↗
One farm in Cyprus: low bacterial counts; sequencing showed mainly Gram-negative environmental bacteria, with lactic acid bacteria at around 2 %.
Papademas et al. · J. Dairy Sci. · Original source ↗
Fermented donkey milk before and after simulated digestion: peptides, ACE inhibition and antioxidant and antimicrobial activity measured in the laboratory.
Aspri et al. · Food Chemistry · Original source ↗
Fermented with two lactic acid bacteria: 679 peptides, somewhat stronger radical scavenging in the test tube. In the virus test, unfermented milk also showed inhibition.
Cirrincione et al. · Molecules · Original source ↗
16 jennies, 3 months: a trace element supplement did not change milk yield or milk components. Protein fell and milk sugar rose over the course of lactation.
Fantuz et al. · Animals · Original source ↗
153 samples from five farms: protein fell from 3.09 to 1.14 %, milk sugar rose. Lactoferrin increased, lysozyme decreased.
Plotuna et al. · Foods · Original source ↗
36 pooled samples over 12 months: calcium, phosphorus, potassium and copper varied seasonally. Lead was above the EU limit in two April samples.
Ljubojević Pelić et al. · Animals · Original source ↗
30 jennies, 8 weeks: concentrate feed gave more milk and better protein digestibility. Only with the herbal supplement did the bacterial count of the milk fall.
Plotuna et al. · Animals · Original source ↗
42 raw milk samples, many heating variants: the cow’s milk limit does not detect insufficiently heated donkey milk. The authors propose 76 °C for 15 s as a reference.
Marino et al. · Foods · Original source ↗
40 jennies, three months of lactation: fat content fell from 0.70 to 0.55 %, saturated fatty acids increased slightly. Redox parameters shifted.
Naydenova et al. · Foods · Original source ↗
12 jennies, 5 weeks: with fermented Codonopsis residue they gave around 32 % more milk, while composition stayed the same. Foals gained more weight.
Zhou et al. · Front. Vet. Sci. · Original source ↗
6 jennies, 7 time points: 1,552 fat molecules recorded. Colostrum was richer in phospho- and sphingolipids, mature milk in glycerolipids.
Du et al. · Front. Nutr. · Original source ↗
3 jennies: 505 proteins recorded in milk vesicles. Colostrum contained twice as many vesicles, and 164 proteins were more abundant there.
Caira et al. · Metabolites · Original source ↗
Proteomic comparison: 241 whey proteins differed between donkey and human milk, 365 compared with cow’s milk. J chain 18 times more abundant in donkey milk than in human milk.
Zhang et al. · Front. Nutr. · Original source ↗
Proteomics of the fat droplet envelope: of 454 proteins, 293 differed between donkey and human milk, 204 of them more abundant in donkey milk.
Zhang et al. · Front. Nutr. · Original source ↗
14 jennies, 10 weeks: an extract of Ordos mugwort increased milk fat and the proportion of unsaturated fatty acids; the gut flora shifted.
Meng et al. · Front. Microbiol. · Original source ↗
12 jennies with foals, 23 days: a yeast polysaccharide did not change milk yield or milk composition. IgA and IgG rose in the jennies’ blood.
Huang et al. · Front. Microbiol. · Original source ↗
Roll cheese with 0, 10 and 20 % donkey milk: less fat, the same bacterial counts, similarly good ratings. Tasters recognised the difference in the triangle test.
Vidaković Knežević et al. · Foods · Original source ↗
18 pooled samples over nine months on gel: in donkey milk, caseins appeared only from month 3, lactoferrin and albumin throughout. No statement about infants.
Georgieva et al. · Foods · Original source ↗
28 jennies, one lactation: according to the model, North African jennies gave 188.66 kg of milk with 1.45 g/100 g protein, Masri jennies 163.42 kg with more fat (1.16 g/100 g). Milk analysis.
Aroua et al. · Animals · Original source ↗
81 samples from 9 Halari jennies: 0.86 % fat, 2.03 % protein, 5.75 % lactose, calcium 61.25 mg/100 mL. Descriptive milk analysis with inconsistencies in vitamins and amino acids.
Garhwal et al. · Foods · Original source ↗
Proteomics on milk from 84 Dezhou jennies (day 1 to 180): 805 proteins; compared with colostrum, 226 were more abundant in mature milk and 219 less abundant. Relative amounts.
Zhou et al. · Foods · Original source ↗
20 Ragusana jennies, 10 with a 30-day probiotic supplement: 12 of 76 milk metabolites shifted, ALP measurable in 6 of 10 animals (0.97–11.73 U/mL). Paper with inconsistencies.
Laus et al. · Metabolites · Original source ↗
Raw, UV-C-treated or pasteurised, then freeze-dried: protein the same (15.3–15.7 %), degree of hydrolysis after laboratory digestion 37 % (raw, UV-C) vs. 44 % (pasteurised).
Tedeschi et al. · LWT · Original source ↗
Protein 1.4 → 2.72 % after ultrafiltration; only this milk coagulated with rennet. Highest cheese yield with starter culture (10.97 %), statistically equal to plain UF (9.60 %).
Natrella et al. · Int. Dairy J. · Original source ↗
Dessert made from fermented donkey milk with oats and oil: texture adjustable via the oat-to-oil ratio. Over 30 days at 4 °C, polyphenols (1.71 → 0.66 mg/g) and firmness decreased.
Antonino et al. · Appl. Food Res. · Original source ↗
Nine short-time heat treatments (75–92 °C, 3–12 s): lysozyme fell from 1.17 to 0.94–0.46 g/L, measured B vitamins without a significant decrease. Laboratory values, not a recommendation.
Matera et al. · LWT · Original source ↗
Casein and whey digested in the test tube: whey was broken down more (donkey whey 73.41 %), donkey casein yielded the most peptides (4.94 mg/mL). Chemical laboratory tests only.
Akan · J. Food Sci. Technol. · Original source ↗
One year of milk analysis in Amiata jennies: lysozyme activity was highest in the 3rd month of lactation (1,842.21 U/mL); no significant difference between raw and pasteurised at 65 °C.
Martini et al. · Int. Dairy J. · Original source ↗
Whey protein from donkey milk broken down with four enzymes: degree of hydrolysis up to 39.46 %, solubility from 47.74 % to close to 80 %, smaller particles. Laboratory measurements only.
Wei et al. · Food Chem. X · Original source ↗
80 adults tasted caciotta made from cow’s milk with and without 5 % donkey milk. Overall impression after 10 days 7.29 vs. 6.71, after 45 days 6.87 vs. 7.56 (9-point scale).
Cosentino et al. · Animals · Original source ↗
20 jennies over ten months: daily milk fell from 1,535 to 869 mL, casein decreased with each foaling, more milk in spring–summer than in autumn–winter.
Salari et al. · Animals · Original source ↗
14 Pêga jennies, seven measuring days: after 2 hours of separation, few changes in behaviour and salivary cortisol. On average 566.4 mL of milk per hand milking.
Farias et al. · Animals · Original source ↗
24 jennies, three energy levels: calcium fell from 1,351 mg/L at birth to 889 mg/L after 14 days. Medium feed energy was mostly associated with the highest mineral values.
Hui et al. · Animals · Original source ↗
9 jennies, 62 samples: daily milk around 2 kg up to day 189, then a decline to 0.13 kg. Protein from 1.59 to 1.28 g/100 g; calcium and phosphorus fell late in lactation.
Malacarne et al. · Animals · Original source ↗
Fresh donkey milk and donkey milk reconstituted from freeze-dried powder from 18 Ragusana jennies compared: protein 1.62 vs. 1.61 %, energy 38.2 vs. 39.0 kcal/100 ml – no significant differences.
Polidori et al. · Food Science and Nutrition Technology · Original source ↗
27 jennies received maize straw, wheat husks or wheat straw for six weeks. 43 volatile compounds were detected in the milk; 23 of them distinguished the feed groups statistically.
Du et al. · Frontiers in Nutrition · Original source ↗
Serbian fresh cheese from cow’s milk with 30 % donkey milk: coagulation took 2 h instead of 30 min, yield fell from 1.37 to 1.10 kg, and protein content rose from 13.45 to 15.75 %.
Ljubojević Pelić et al. · Animals · Original source ↗
15 Nordestino jennies: after 4 h of udder fill time, an average of 312.38 g of milk was obtained per milking, after 2 h 188.76 g. Milking three times a day gave 724 g per day (median) compared with 131 g with milking once a day.
Santana et al. · Veterinary and Animal Science · Original source ↗
In six Dezhou jennies, the fat content of the milk fell from 0.92 % on the first day to 0.09–0.20 % from day 90, while lactose rose from 3.94 % to up to 6.85 %. Hormones and blood values of the animals were recorded in parallel.
Zhou et al. · Veterinary Medicine and Science · Original source ↗
Laboratory comparison of milk on day 3 after birth in four species: in donkey colostrum (2.29 % protein, 1.10 % fat), 303 proteins in the soluble fraction changed after 72 °C for 15 s, but none of the 38 fatty acids measured.
Zhao et al. · Foods · Original source ↗
In milk from Dezhou donkeys, 1774 lipids and 35 volatile compounds were recorded. Between colostrum (day 3) and mature milk (month 3), 233 lipids and 11 volatile compounds differed significantly.
Li et al. · Food Chemistry: X · Original source ↗
16 lactating Dezhou jennies received feed with or without 0.5 g/kg mannan oligosaccharides for 60 days. Milk fat (0.65 vs. 0.87 %) and milk protein (1.69 vs. 1.79 %) did not differ significantly; the foals tended to gain more weight (25.63 vs. 31.75 kg).
Li et al. · Microorganisms · Original source ↗
Twelve jennies on the same feeding were grouped by milk fat (0.42 % vs. 0.13 %). Only 11 lipids differed in the blood, but 131 in the milk – mainly triglycerides.
Zhang et al. · Veterinary and Animal Science · Original source ↗
In donkey and bovine colostrum, 206 polar lipids from 11 subclasses were recorded; 141 differed markedly. Sphingomyelin made up 70.7 % of the polar lipids in donkey colostrum and 50.9 % in bovine colostrum.
Zhang et al. · Food Chemistry: X · Original source ↗
In vesicles from the milk of three Dezhou jennies, 292 miRNAs were detected, 48 of them possibly novel. 73 mature miRNA sequences were also found in data sets on human and cow’s milk.
Yan et al. · Food Chemistry: Molecular Sciences · Original source ↗
In 24 Dezhou jennies, the medium energy level (12.4 MJ/kg) gave the highest milk yield obtained, at 0.93 kg/d (high: 0.86, low: 0.75 kg/d). Between the low and medium levels, 176 milk metabolites and 87 whey proteins differed.
Zhao et al. · Antioxidants · Original source ↗
With 0.3 mg selenium per kg of feed dry matter, the extrapolated milk yield rose from 3.26 to 3.84 kg/d; the fat, protein and lactose contents of the milk stayed the same. The share of saturated fatty acids fell from 56.74 % to 54.06 % (0.5 mg: 50.90 %).
Meng et al. · Animals · Original source ↗
With 5 g of methionine per day, the milk yield obtained was 1.72 kg compared with 1.15 kg without supplement; 15 g/d showed no effect on yield (1.14 kg). 568 metabolites were recorded in the milk, 41 of which differed between the groups.
Huang et al. · Animals · Original source ↗
28 jennies received soybean and linseed oil in four mixing ratios for four weeks. 1167 lipids and 406 volatile compounds were recorded in the milk; 64 lipids and 89 volatile compounds differed between the groups.
Ren et al. · Food Science of Animal Resources · Original source ↗
27 jennies received soybean, linseed or palm oil in their feed for four weeks. 725 lipids and 391 volatile compounds were recorded in the milk; fat content in the linseed oil group was 0.07 % compared with 0.25–0.26 %.
Ren et al. · Food Science & Nutrition · Original source ↗
In 28 lactating dairy jennies, the teats were measured by ultrasound: the teat canal was on average 1.85 cm long and 0.41–0.46 cm wide. The width of the teat canal increased with the month of lactation (r = 0.32 and 0.54).
Accorroni et al. · Veterinary Sciences · Original source ↗
In 112 Xinjiang jennies, 236 gene variants were found that were statistically associated with milk yield or fat, protein or lactose content. In a validation group of 82 animals, daily milk yield differed by genotype for eight variants (e.g. 2.67 vs. 1.57 kg).
Fang et al. · International Journal of Molecular Sciences · Original source ↗
A chip sensor detected cow’s milk in donkey, sheep’s and goat’s milk from 0.05 % (v/v) in 12 minutes. Recovery was 92.0–105 %, and measurement variation 4.0 % (same day) and 6.5 % (different days).
Kourti et al. · Sensors · Original source ↗
Along the processing chain of a farm dairy, protein content was 13.5 g/kg in raw milk and 16.5 g/kg in reconstituted powder. In total, 1505 metabolites were recorded; 354 differed between the four forms of milk.
Zhang et al. · Food Chemistry: X · Original source ↗
From 29 public sequencing data sets, including 4 from donkey milk, 1549 conserved miRNAs were recorded. miR-148a-3p made up the largest share in all four types of milk, on average about 21.5 %.
Zoziuk et al. · International Journal of Molecular Sciences · Original source ↗
683 lipids were recorded in milk from nine species. In the triglycerides of donkey milk, polyunsaturated fatty acids were 26.90 mol% (human milk 29.96, cow’s milk 3.17 mol%).
Chen et al. · npj Science of Food · Original source ↗
640 lipids were recorded in raw milk from eight species. Donkey milk had the lowest fat content at 0.52 %; with the same sample volume, 548 of the 640 lipids were lower than in Holstein milk.
Wei et al. · Foods · Original source ↗
Five milk powders made in the laboratory were compared with two methods: 55 and 86 volatile compounds were recorded. 58 of the 86 compounds were found in the donkey milk powder, 25 of them only there, including several ethyl esters.
Zhou et al. · Food Chemistry: X · Original source ↗
61 volatile compounds were recorded in cow, donkey, camel and horse milk powders from Xinjiang. Eight of them reliably distinguished donkey and cow milk powder, for example 3-hydroxy-2-butanone (0.55 vs. 5.60 %).
Gou et al. · Foods · Original source ↗
In milk from 49 jennies from Lazio, miRNAs made up 7.74 % of the mapped small RNAs, compared with only 0.87–1.57 % in buffalo, goat and sheep. Only four miRNAs occurred in all four types of milk.
Cendron et al. · BMC Genomics · Original source ↗
Cow’s, goat’s, buffalo and donkey milk were treated with ultrasound (200–1000 W) or heat (65–100 °C). In donkey milk, the β-sheet share of the proteins rose from 43.89 % to up to 48.33 %; its rennet gel was too weak for a texture measurement.
Hou et al. · Ultrasonics Sonochemistry · Original source ↗
DNA was obtained from 10 mL of milk each from ten Halari jennies; a 235 bp section of the κ-casein gene (CSN3) was detected in all samples. The sequence matched the reference by 99 %.
Singh et al. · Heliyon · Original source ↗
In mammary gland tissue of Liaoxi jennies, 3144 mRNAs and 293 miRNAs differed between lactation (day 60) and the dry period. Genes for caseins were around 17 to 18 times more active during lactation.
Fei et al. · Genes · Original source ↗
In eight donkey milk samples, odd- and branched-chain fatty acids amounted to 2.36 g/100 g fatty acids – less than in ruminants (3.33–4.51), more than in human milk (1.34). Undecanoic acid (C11:0) reached 0.895 g/100 g.
Carta et al. · Foods · Original source ↗
In 24 Dezhou jennies, the extrapolated milk yield with 14.2 % or 15.3 % crude protein in the feed was 3.27 and 3.41 kg/d, and with 13.1 % 2.80 kg/d. Milk protein content rose slightly from 1.82 to 1.87–1.89 %.
Yue et al. · Animals · Original source ↗
Milk from eight Zamorano-Leonés jennies contained 1.68 % protein, 0.56 % fat and 6.6 % lactose at 155 kJ/100 g. Polyunsaturated fatty acids made up 27.06 % of the fatty acids, oleic acid 25.12 %.
Albertos et al. · Frontiers in Nutrition · Original source ↗
In nine Amiata jennies, the IgG content of colostrum fell from 1.95 log10 mg/mL directly after birth (around 89 mg/mL) to 1.40 log10 mg/mL after 24 hours (around 25 mg/mL). In the foals’ blood, IgG rose over the same period from 8.97 to 14.91 mg/mL.
Turini et al. · Heliyon · Original source ↗
With microbial rennet, donkey milk could be coagulated in around 9 minutes; the fresh cheese achieved a yield of 7.21 % with 25.87 % protein and 2.90 % fat. The comparison cheese from cow’s milk reached a yield of 23.52 %.
Faccia et al. · Foods · Original source ↗
In nine Martina Franca jennies, milk fat content fell from 0.72 % (day 15) to 0.42 % (day 210) and energy value from 1857.3 to 1645.6 kJ/kg. The share of saturated fatty acids fell from 62.88 % to 41.54 % on day 180.
Martemucci et al. · Lipids in Health and Disease · Original source ↗
633 proteins were recorded in six pasteurised donkey milk samples from three Greek regions. Only 19 of them were assigned to donkey or horse sequences; 505 are based on a single unique peptide.
Proikakis et al. · Data in Brief · Original source ↗
In milk from Martina Franca jennies, lysozyme fell from 1.34 mg/mL (day 60) to 0.76 mg/mL (day 190); β-lactoglobulin averaged 3.75 mg/mL and lactoferrin 0.080 g/L.
Polidori et al. · Foods · Original source ↗
The largest research field – in mice and rats. Not transferable to humans.
Mice with small-intestinal inflammation: in the study, those given raw donkey milk showed less tissue damage and fewer inflammatory markers; heated milk showed no comparable effect.
Yvon et al. · Eur. J. Nutr. · Original source ↗
Donkey milk lysozyme in a mouse model of colitis: researchers observed less weight loss, more barrier proteins and fewer inflammatory markers.
Jiang et al. · J. Funct. Foods · Original source ↗
Mice under stress: in the study, donkey milk and a lysozyme fraction attenuated intestinal changes, and bifidobacteria returned to control levels.
Yvon et al. · Gastroenterology · Original source ↗
In the study, 72 °C for 2 minutes retained the most lysozyme. In the mouse stress model, this milk affected intestinal permeability similarly to raw milk.
Yvon et al. · J. Dairy Sci. · Original source ↗
Exosomes from donkey milk carrying 1,212 microRNAs in a mouse model of colitis: two microRNAs dampened inflammatory signalling pathways.
Zhong et al. · J. Food Sci. · Original source ↗
The same mouse study as study 16, here focusing on the Paneth cells: lysozyme, defensins and selected gut bacteria in the ileitis model.
Yvon et al. · Eur. J. Nutr. · Original source ↗
The same work as study 19, here focusing on the enzyme: how much lysozyme activity remains after thermisation – and what was measured in the mouse stress model.
Yvon et al. · J. Dairy Sci. · Original source ↗
Donkey milk powder in a mouse model of colitis: TLR4/MyD88/NF-κB less active, barrier markers ZO-1 and MUC2 higher.
Yang et al. · Food Science & Nutrition · Original source ↗
Heated, sonicated or UV-treated whey proteins in a mouse model of colitis: in the study, the unheated proteins gave better results.
Cai et al. · J. Dairy Sci. · Original source ↗
Mice, 28 days of donkey milk: gut flora and immune parameters changed, including more Lactobacillus.
Meng et al. · Frontiers in Veterinary Science · Original source ↗
Rats given donkey milk for ten days before gastric mucosal injury: in the study, less damage, less MDA and more glutathione were observed.
Sami et al. · Vet. Med. Sci. · Original source ↗
8 oligosaccharides detected, mainly 3′- and 6′-sialyllactose. In the cell experiment, an extract slowed the division of intestinal cells and promoted their maturation.
Wang et al. · Food & Function · Original source ↗
Two animal studies – first indications that need to be independently replicated.
Rats with type 2 diabetes, four weeks of donkey milk powder: in the study, blood glucose was around 36 % below the comparison group, HOMA-IR 75 instead of 112.
Li et al. · J. Ethnopharmacol. · Original source ↗
Rats given human, donkey or cow’s milk: in the study, the human and donkey milk groups showed lower inflammatory markers and differences in glucose and lipid metabolism.
Trinchese et al. · Frontiers in Physiology · Original source ↗
Rats drank cow’s, donkey or human milk for 4 weeks (82 kJ/day each). In the heart tissue of the donkey milk group, DHA was 11,132.58 nmol/g (control 8,938.36). Laboratory measurements.
Trinchese et al. · Antioxidants · Original source ↗
Mice received donkey milk or a sugar fraction from it for 30 days. Swimming time to exhaustion: control 51.50 min, high donkey milk dose 116.67 min. Three authors employed by a company.
Wang et al. · Foods · Original source ↗
Disease model: rats on high-fat feed received donkey milk or selenium donkey milk (selenium 0.082 instead of 0.024 mg/kg) for 8 weeks. Values mostly shown only as graphs, several contradictions.
Zhong et al. · Foods · Original source ↗
Cell experiments and one study on mouse skin. No statement about drinking.
Donkey milk powder on UVB-irradiated cells and mouse skin: in the study, melanin and tyrosinase decreased and the barrier protein filaggrin increased.
Li et al. · Frontiers in Nutrition · Original source ↗
Mouse connective tissue cells with four types of milk: donkey milk activated Erk three- to fourfold and dampened NF-κB.
Kocić et al. · Oxid Med Cell Longev · Original source ↗
Formulation trial: from 1, 2 and 5 per cent donkey milk solutions, about 83.5 % of the protein was assigned to nanoliposomes, from 10 % solution 88.9 %. No skin or cell tests.
Kocić et al. · Acta Fac. Med. Naiss. · Original source ↗
Experiments in mice, piglets and rabbits and a small pilot study in hospital.
Mouse model of asthma: in the study, airway hyperresponsiveness was reduced, and eosinophils, IgE and the mucus protein MUC5AC were lower.
Lu et al. · J. Funct. Foods · Original source ↗
48 piglets around weaning given donkey or cow’s milk: inflammatory gene IL1B 8.11 compared with 15.55.
de Souza Farias et al. · Frontiers in Immunology · Original source ↗
60 adults with pneumonia, 30 of whom also received donkey milk: in the study, individual inflammation values differed between the groups. Pilot study without placebo.
Kolarov et al. · Antibiotics · Original source ↗
45 rabbits after a vaccine stimulus: raw milk and powder showed very similar measurement patterns, statistically equivalent for lysozyme in the blood.
Plotuna et al. · Foods · Original source ↗
Basic research: it describes processes in the laboratory – not the effect of a food.
Mouse tumour model with donkey milk kefir or donkey milk: plain donkey milk did not reduce tumour growth.
Esener et al. · Biotech. Histochem. · Original source ↗
4T1 mouse tumour model: donkey milk changed cell death and immune signals, and messenger substances shifted towards Th1.
Zhang et al. · J. Funct. Foods · Original source ↗
Milk fractions on lung cancer cells and mouse immune cells: a whey protein fraction showed the clearest effect.
Mao et al. · Int. Dairy J. · Original source ↗
Normal and lung cancer cells responded differently; no damage to genetic material in normal cells at 50–400 µg/ml.
Akça et al. · Czech J. Food Sci. · Original source ↗
Raw milk from 31 jennies on gastric cancer cells: effects depended on the proportion of milk; composition remained stable over the lactation.
Mazzone et al. · Biology · Original source ↗
Milk digested in the laboratory on Caco-2 cells: donkey and sheep’s milk triggered different gene activity patterns. No measurable cell damage. Cell culture only.
Bogdanović et al. · Dairy · Original source ↗
Clinical studies with children under medical supervision and one laboratory study. The results apply only to the group examined.
46 selected children with confirmed cow’s milk allergy, under medical supervision: in the study, 38 (82.6 %) tolerated and accepted donkey milk.
Monti et al. · Pediatr. Allergy Immunol. · Original source ↗
92 children with cow’s milk allergy, under medical supervision: in the study, 83 (90.2 %) tolerated donkey milk. Beta-lactoglobulin was cross-reactive.
Monti et al. · J Biol Regul Homeost Agents · Original source ↗
In the study, 77 of 78 children tolerated the medically supervised challenge; 22 grew normally over six months on a supplemented diet.
Sarti et al. · Ital. J. Pediatr. · Original source ↗
Sera from three adults with cow’s milk allergy reacted to cow’s and in part goat’s milk proteins, but not to donkey milk proteins. Very small number of cases.
Vincenzetti et al. · Veterinaria Italiana · Original source ↗
25 children with confirmed cow’s milk allergy were tested under medical supervision: according to the study, 24 (96 %) tolerated donkey milk in an open food challenge; follow-up over 4–6 months.
Tesse et al. · Italian Journal of Pediatrics · Original source ↗
The FortiLat series and one laboratory experiment – always with a medical breast milk fortifier, not with donkey milk for drinking.
156 preterm infants, fortifier based on donkey or cow’s milk: no difference in the primary outcome; in the study, fewer episodes of vomiting and feeding intolerance were recorded.
Bertino et al. · J Pediatr Gastroenterol Nutr · Original source ↗
122 children at 18 months: weight, length and head circumference without a significant difference between the groups.
Peila et al. · Nutrients · Original source ↗
Follow-up of 103 children: the developmental quotient was almost the same in both groups (median 101.2 vs. 101.6).
Peila et al. · Nutrients · Original source ↗
Sub-analysis of 10 infants: 2.02 reflux events per hour with the donkey milk fortifier, 4.82 in the comparison group (median).
Cresi et al. · Nutrients · Original source ↗
Follow-up survey at 6 to 8 years: no difference between donkey and cow’s milk fortifier in any of the nine allergic endpoints.
Peila et al. · Breastfeed. Med. · Original source ↗
Simulated preterm infant digestion in the laboratory: with the donkey milk fortifier, different peptides formed than with the cow’s milk fortifier. Note the patent conflict of interest.
Giribaldi et al. · Food Chemistry · Original source ↗
Not studies but historical texts from antiquity to the early 20th century – each with a verifiable original passage and critical assessment.
The Greek medical treatise from around 400 BC mentions donkey milk in more than ten chapters – mostly as a drink given after a purgative, often alternating with whey.
Hippocratic Corpus (anonymous author, Cnidian tradition) · De morbis II (Diseases, Book 2) · Original source ↗
In the 4th century BC, Aristotle ranked types of milk by their thickness: camel’s milk the thinnest, then human and donkey milk, cow’s milk the thickest. Mare’s and donkey milk, he wrote, were mixed into Phrygian cheese.
Aristotle · Historia animalium (History of Animals) · Original source ↗
In 37 BC, the Roman scholar Varro ranked the types of milk: sheep’s milk was the most nourishing, mare’s milk the most strongly purging – with donkey milk in second place.
Marcus Terentius Varro · Res rusticae (On Agriculture) · Original source ↗
No ancient text reports that Cleopatra bathed in donkey milk. The sources name Poppaea Sabina, Nero’s wife, around 95 years after Cleopatra’s death.
Cross-check of ancient sources (Pliny the Elder, Cassius Dio) and research literature · Fact check: did Cleopatra bathe in donkey milk? · Original source ↗
Around AD 77, Pliny reports that Nero’s wife Poppaea took 500 she-asses with her and bathed in their milk. He also reproduces the opinion that donkey milk smooths the skin of the face.
Pliny the Elder (Gaius Plinius Secundus) · Naturalis historia · Original source ↗
In Book 28 of his Natural History, Pliny calls donkey milk “the most efficacious” milk and lists around a dozen uses attributed to it at the time – as a reproduction of other people’s doctrines.
Pliny the Elder (Gaius Plinius Secundus) · Naturalis historia · Original source ↗
Around AD 116, the satirist Juvenal ridicules a rich Roman woman who smells of “Poppaea ointments” and tends her skin with donkey milk – she would take she-asses with her even into exile at the North Pole.
Decimus Iunius Iuvenalis (Juvenal) · Saturae, Satire 6 · Original source ↗
Around AD 175, the physician Galen advised older people to drink goat’s and donkey milk alternately: donkey milk, he said, was thin and rich in whey, goat’s milk of medium consistency.
Galen of Pergamon (Galen) · De sanitate tuenda (On the Preservation of Health) · Original source ↗
In the early 3rd century, the historian Cassius Dio reports that Poppaea had 500 she-asses milked every day in order to bathe in their milk – as an example of her luxury.
Cassius Dio (Lucius Cassius Dio) · Roman History (Rhomaika) · Original source ↗
In two letters to her daughter, the Marquise de Sévigné mentions donkey milk being taken as a regimen in her Parisian circle of acquaintances – drunk “in the evening and in the morning”. The letters show how commonplace donkey milk was in the France of Louis XIV.
Marie de Rabutin-Chantal, Marquise de Sévigné (1626–1696) · Lettres (letters to her daughter, the Comtesse de Grignan), Monmerqué edition · Original source ↗
In the fourth volume of his Natural History in 1753, Buffon states that the use of donkey milk extends “from the Greeks down to us” and describes how a milk-producing she-ass should be selected and fed.
Georges-Louis Leclerc, Comte de Buffon (1707–1788) · Histoire naturelle, générale et particulière, Tome IV · Original source ↗
The 50-page pamphlet translates the Halle physician Friedrich Hoffmann into English. It contains simple comparative experiments: from 12 ounces of donkey milk Hoffmann obtained barely 2 drachms of curd, from cow’s milk 10 drachms.
Friedrich Hoffmann (1660–1742), English translation by an anonymous translator · A Treatise of the Extraordinary Virtues and Effects of Asses Milk · Original source ↗
In 1777, the extensive German Oeconomische Encyclopädie describes donkey milk as “lean” compared with cow’s milk and closer to human milk – and reproduces a recipe typical of its time for “artificial donkey milk”.
Johann Georg Krünitz (1728–1796) · Oeconomische Encyclopädie, vol. 11 · Original source ↗
In 1789, Robert Hammond of Knightsbridge defended himself in a handbill against competitors who falsely claimed to be successors of his house. His house, he wrote, had been supplying customers with donkey milk for over 60 years.
Robert Hammond, donkey milk dealer in Knightsbridge · Handbill “To the Nobility, Gentry, and Others” (printed by H. Reynell, Piccadilly) · Original source ↗
The two Parisian chemists compared the milk of women, cows, goats, she-asses, ewes and mares. In their ranking table, donkey milk comes first for whey content and second for milk sugar.
Antoine-Augustin Parmentier (1737–1813) and Nicolas Deyeux (1745–1837) · Précis d’expériences et observations sur les différentes espèces de lait · Original source ↗
An often-told story: a Jewish physician from Constantinople is said to have prescribed donkey milk for King Francis I. The oldest versions we were able to check come from 19th-century reference works – not from the king’s time.
François Noël (1755–1841) et al.; Pierre Larousse (1817–1875) · Nouveau dictionnaire des origines (1827); Grand dictionnaire universel du XIXe siècle, vol. 1 (1866) · Original source ↗
Documents from authorities, parliaments and international organisations – summarised factually, with a link to the original.
In Table 3.3, the 2013 FAO book lists donkey milk with an average of 90.8 g water, 1.6 g protein, 0.7 g fat and 6.4 g lactose per 100 g. Donkeys are counted there among the “minor dairy species”.
Food and Agriculture Organization of the United Nations (FAO); technical eds. Ellen Muehlhoff, Anthony Bennett, Deirdre McMahon; Chapter 3: Ramani Wijesinha-Bettoni, Barbara Burlingame · Milk and dairy products in human nutrition (ISBN 978-92-5-107863-1) · Original source ↗
The EU market regulation reserves the term “milk” for the product of normal mammary secretion. For milk that does not come from cows, the animal species must be stated – hence “donkey milk”.
European Parliament and Council of the European Union · Regulation (EU) No 1308/2013 establishing a common organisation of the markets in agricultural products (OJ L 347, 20.12.2013, p. 671) · Original source ↗
Among other things, the 2017 guidelines recommend starting milking 20 to 90 days after the foal’s birth and not extending lactation beyond 270 days.
The Donkey Sanctuary (United Kingdom) and Università degli Studi di Milano; Francesca Dai, Emanuela Dalla Costa, Michela Minero, Faith Burden, Andrew Judge · Dairy donkeys: Good animal management practices for donkey milk production · Original source ↗
In December 2022, Indian research institutes met with ILRI to discuss donkey milk. The director of the equine research centre called for the approval of donkey milk by the food authority FSSAI to be accelerated.
Indian Council of Agricultural Research (ICAR), source: ICAR-National Research Centre on Camel, Bikaner · Stakeholders Meet on “Improvement of Donkey and Non-Bovine Milk” organized at ICAR-NRCC · Original source ↗
The official breed sheet of the Emilia-Romagna region lists 1,152 registered Asino Romagnolo donkeys on 350 farms and names milk production as a new use alongside carrying, draught and mule breeding.
Regione Emilia-Romagna (regional register of genetic resources under Regional Law L.R. 1/2008) · Scheda tecnica per l’iscrizione al repertorio: Asino Romagnolo RER A020 (Revisionato), Allegato 10, Bollettino Ufficiale della Regione Emilia-Romagna n. 381 of 13.12.2024 (Parte seconda) · Original source ↗
In a parliamentary answer of 17.12.2024, the Indian government names two ICAR-NRCE research projects on donkey milk and a support programme with grants of up to 50 lakh rupees that also covers donkeys.
Government of India, Ministry of Fisheries, Animal Husbandry and Dairying (Department of Animal Husbandry and Dairying); answer by Minister Rajiv Ranjan Singh alias Lalan Singh · Lok Sabha, Unstarred Question No. 3580 “Non-Bovine Milk”, answered on 17.12.2024 · Original source ↗
China’s 2024 milk powder standard names donkey milk powder for the first time: at least 2.5 g fat per 100 g and protein of at least 18 % of the non-fat solids. The animal species must be stated.
National Health Commission (NHC) and State Administration for Market Regulation (SAMR) of the People’s Republic of China · National Food Safety Standard GB 19644-2024 “Milk Powder and Modified Milk Powder” (乳粉和调制乳粉), in force since 8.2.2025 · Original source ↗
No study matches this selection.
We render each publication in our own words based on the original or the official abstract. The linked source is authoritative.
Summaries as of: October 2026
Research portrait
Alongside the published studies, we keep an eye on who is actively researching donkey milk today. Prof. Hans-Christian Siebert is not interested in the milk as a whole but in individual building blocks: proteins that recognise sugars, and tiny vesicles that can transport molecules.
Physicist and veterinarian · co-founder of the RI-B-NT Research Institute of Bioinformatics and Nanotechnology, Kiel
His institute was founded in 2011 at the Kiel Innovation and Technology Centre and is now part of an international research network for nanomedicine. He combines biophysics, biochemistry and computer models to understand how proteins interact with sugar chains on cell surfaces – a field known as glycobiology.
What his group researches
Lysozyme is known for breaking down bacterial cell walls. The group is investigating a second property: like a so-called lectin, lysozyme can recognise and bind certain sugar structures.
For a 2024 study, freeze-dried lysozyme from donkey milk was redissolved and examined by NMR spectroscopy.
Zhang R. et al. 2017 · Zhang N. et al. 2024
Sialidases are enzymes that cleave off certain sugar building blocks – sialic acids. Such structures play a role in communication between cells.
In the group’s milk comparisons, sialidases were found only in donkey milk and in llama and dog colostrum.
Schauer & Kamerling 2018 · RI-B-NT
Exosomes are tiny vesicles in milk enclosed by a lipid envelope. Their envelope can protect the contents from breakdown.
The group is working on making exosomes from donkey milk usable as standardisable vesicles for nanomedical applications – a research goal, not yet an application.
Liu Y. et al. 2025 · RI-B-NT
It turned out that lysozymes and sialidases as well as sialylated oligosaccharides occur in high concentrations in the exosomes of donkey milk.
Prof. Hans-Christian Siebert
Using natural and cell-produced donkey milk, we are working on making the exosomes usable as nanomedical tools in the form of standardisable and scalable vesicles.
Prof. Hans-Christian Siebert
Seven types of milk compared
The group compared the constituents of donkey milk and donkey colostrum with milk and colostrum from other farm and domestic animals – using mass spectrometry, which makes even the smallest amounts of individual molecules visible.
This is basic research on individual constituents of donkey milk. The work described makes no statement about the effect of donkey milk as a food and was not carried out with DON DONKEY'S products. The quotes reflect Prof. Siebert’s scientific view.
