Here, we studied the associations of companion dogs’ early life diet style with CE incidence in later life. Our results showed that feeding a non-processed or minimally processed meat-based diet to puppies and adolescent dogs at 2 to 18 months associated significantly with decreased CE incidence in adulthood. On the contrary, feeding an ultra-processed dry dog food (kibble) based diet associated significantly with increased CE incidence in adulthood. These findings support our previous observation that the consumption of a NPMD during PU may protect against CE later in life and that consumption of an UPCD increases CE incidence later in life12. These two diets differ essentially in their degree of processing28 and protein:fat:carbohydrate (PFC) macronutrient ratio. In NPMD the macronutrient ratio is typically 45:50:0–10 percent dry matter whereas in UPCD it is 16–38:6–18:40–6029.
The absence of a significant protective effect on CE later in life when feeding a home-cooked diet that is composed partly of similar, but cooked, food items as the non-processed meat-based diet, suggests that the lack of processing of the ingredients may be of importance. Dry dog food is ultra-processed by heat treatment, rendering, milling, and/or extrusion and contains food additives such as emulsifiers, colouring agents and palatability enhancers28,30,31,32,33. In contrast, meat-based non-processed foods for dogs are composed of fresh ingredients which may be chopped, mixed, and frozen and the only additives they might contain are minerals and vitamins if needed for balancing34,35. Thermal processing of foods containing carbohydrates and protein produces Maillard reaction products including advanced glycation end products (AGEs), which are immunomodulatory and may increase the prevalence of diet-related chronic inflammatory states in the gut36,37 although the current literature on the topic is partly controversial38. In addition, although the number of feed additives in ultra-processed carbohydrate-based dog food is vast, very little is known about their single or synergistic effects on gastrointestinal health33,39,40. Analogously, Western diet containing high amounts of ultra-processed foods and refined carbohydrates has been suggested as a potential explanation of increased prevalence of human IBD in the industrialized world24.
Domesticated dogs (Canis familiaris) are members of the order Carnivora, but can utilise non-animal foods such as plant material and carbohydrates in their diet41,42. However, only protein and fat are essential to the dog, they do not have dietary requirements for carbohydrates43. In addition, research indicates that dogs intuitively select a dietary macronutrient composition dominated by protein and fat29,44. Feeding dogs a NPMD is known to affect serum, urine and fecal metabolite concentrations when compared to feeding an UPCD (dry dog food) diet45,46 although the meaning of these changes still requires further research.
Finally, the intestinal bacterial flora plays a fundamental role in the pathogenesis of canine CE and human IBD10,11,16,17. The feeding of a NPMD has been shown to stimulate the growth of a balanced gastrointestinal microbiome in the dog, including an increase of the relative abundances of bacteria associated with protein and fat utilisation, in comparison to dogs fed an UPCD (dry dog food)21,47. Members of these genera also produce butyrate from protein and amino acid fermentation and thus contribute to the intestinal homeostasis. The diet is an essential part of the exposome48 and our observations are in accordance with the biodiversity hypothesis stating that the more microbial exposures in the early life, the more developed immune system in adulthood12,49. This may also explain why eating carcasses outside was negatively associated with CE later in life, contributing to the exposure of diverse microbiota. Fresh ingredients, including vegetables and roots as fibre and phytonutrient sources are part of the NPMD34,35. They contain a variety of microbes of which many may be beneficial for gut health50 as well as indigestible and soluble fibers which act as prebiotics and are fermented in the colon by bacteria to produce gut health promoting short-chain fatty acids51,52,53.
We further studied the associations of the consumed food items with CE incidence in later life. Out of the items belonging to the non-processed meat-based diet style, raw bones and cartilage and berries associated significantly with less CE in adulthood. The protective effect of raw bones and cartilage increased with increasing feeding frequency, being largest when fed daily or almost daily (PU) or a couple of times per week (YO). Eating raw bones and cartilage may have several benefits when fed to puppies and adolescent dogs. Chewing is a behavioural element of dog feeding54,55,56 and thus has been suggested to have a relaxing, stress-relieving effect and it also reduces chewing of unwanted objects when used as enrichment57. Chewing alone, however, does not explain the beneficial effect of bones and cartilage as consuming rawhides, that are chewing items as well, had a predisposing effect on CE incidence later in life as discussed later. Chewing raw bovine femur bones has also been shown effective in removing dental calculus in dogs58,59, but the same effect has been shown for “an extruded, chemically-coated rawhide preparation”60, and therefore cannot explain the beneficial effect of raw bones and cartilage either. While bone is composed primarily of inorganic calcium phosphate61, cartilage contains chondrocytes, collagen, glucosamine and proteoglycan aggregates that are made up of glycosaminoglycans, hyaluronic acid and chondroitin sulfate62,63. Glucosamine is a precursor for synthesis of the glycosaminoglycans and gastrointestinal glycoproteins (mucins) that are essential in mucosal defense64. Glucosamine and glycosaminoglycans modulate microbiota65. Glycosaminoglycans promote intestinal growth and wound repair66, have an effect on the inflammatory mediators and disease activity of IBD patients67, and may improve the intestinal membrane integrity of CE dogs68. Both collagen and collagen hydrolysate treatments have been shown to significantly reduce the mucosal damage score and facilitate faster regeneration of damaged GI mucosa69. Cartilage glycosaminoglycans have been shown to increase the uptake of iron from the diet, avoiding anemia – a symptom very commonly seen in CE70,71.
A whole prey-like diet including ground bone and cartilage as a source of animal fiber containing soluble and insoluble coarse may provide source substrates for gut health promoting short-chain fatty acid production by fermentation72. Mimicking an ancestral, whole-prey diet including bone and cartilage material, is also a possible explanation to the association of eating carcasses outside and less CE later in life.
Our results also indicated a protective role of eating berries against CE later in life. The berries eaten were mainly wild blueberries (Vaccinium myrtillus) which contain a variety of flavonoids, polyphenols, phenolic acids, pyruvic acid, chlorogenic acid, and other compounds which have anti-inflammatory, immunity enhancing and chronic disease preventing properties in humans as reviewed in Ma et al.73. Blueberries have been shown to have a significant antioxidant and anti-inflammatory activity when fed to dogs74. Blueberries were also shown to protect dogs against exercise-induced oxidative stress74. Moreover, in Finland berries are very often picked from the forest and many dogs even eat them directly from the plants in the forest. Therefore, these dogs also get exposed to the diverse forest microbiota which according to the biodiversity hypothesis promotes a healthy microbiome and a balanced immune system4. Eating berries during the berry picking season is a possible explanation why the answer “a couple of times in a year” was significantly different between the control and case dogs. Gray wolves (Canis lupus) consume berries seasonally when available or abundant75 and may even provision pups with them76.
Human meal leftovers and table scraps offered to the puppies and young dogs were found to be significantly associated with less CE later in life. The protective effect increased with feeding frequency, hence the more exposure the dogs had to leftovers, the more protection against CE development there was. Traditionally popular Finnish dishes and meals are composed of fish and meats, vegetables and roots, mushrooms, buttermilk and other fermented milk products, berries, and whole grain products, for example, black rye bread and oatmeal77. The raw meat scraps and the trimmings of the fatty parts might have a similar effect as the NPMD. Moreover, whole grain products as well as roots and vegetables contain indigestible and soluble fibres such as beta-glucans, which have gut health promoting effects51,53,78,79,80. This is a likely explanation for grain products showing a significant association with less CE in adulthood. Another explanation could be the popular Finnish dog food cooked at home; a long time soaked, slow-cooked porridge made of oats, millet, barley, buckwheat, and whole grain rice with vegetables and salt, often served with raw meat and eggs as the protein source (Finnish name Yrjölä porridge). Moreover, healthy humans might share mouth resident beneficial microbes81, like Streptococcus salivarius82, with their dogs, that are transferred to the puppies by contact with table utensils and hands. Furthermore, offering leftovers to dogs is indicative of a more intense human-canine bond, decreasing the stress a puppy experiences and affecting the immune system positively in both dogs and their owners79,83,84,85. Leftovers also decreased chronic gastrointestinal signs after acute gastric dilatation-volvulus surgery in Finnish dogs86.
Organ meats, that were found protective during YO within the home-cooked diet, contain vitamins and micronutrients like vitamins A and B, selenium and coenzyme Q10 which may have a positive impact on gut health87,88,89. It is not clear why blood pancakes were found protective; one possible explanation is blood as a protein and/or a source of iron.
Rawhides fed to puppies a couple of times in a month or more often was associated with an increased risk of CE later in life. The risk increased with increased feeding frequency being highest when rawhides were given more than one time per week during PU. This may be related to low digestibility of rawhides, measured as in vitro dry matter disappearance (DMD), which may pose a risk for gastrointestinal blockage and intolerance especially if the dog tends to swallow large size rawhide pieces which remain intact during gastric and intestinal phases90. However, as rawhides are made from leather industry by-products, it is also possible that the toxic chemicals used for leather processing cause the negative effects on gut health we saw in this study. Unfortunately, there are no scientific publications on analyses of rawhide composition available and clearly requires further research. In 2017, the Food and Drug Administration (FDA) announced a voluntary recall of rawhide chews for dogs due to the use of quaternary ammonia compounds during the rawhide processing91.
There is no conclusive data on the prevalence of CE in dogs, although the disease is frequently diagnosed at animal clinics worldwide (1–17.8%)5. The prevalence of owner-reported CE symptoms was 18–22% in the sample of Finnish FFQ responses. In this study, owners reported onset of CE symptoms at 1.4–1.5 years mean age. Part of the case dogs were also on gastrointestinal or allergy prescription dry dog food diets, commonly suggested by vets for dogs suffering from chronic GI symptoms7already at an early age as could be seen already from the PU and YO datasets. Moreover, eating sticks outside a couple of times per week or more often compared to control dogs during PU and YO, and eating dirt outside more than control dogs during PU, were associated with increased incidence of CE likely because of pica i.e. dogs seeking relief to discomfort or GI symptoms, or stress resulting from the pain92,93. By eating dirt it is also possible that dogs look for micronutrients30,31, prebiotics or lactobacilli, fulvic acid or some other nutrients that they are lacking. Eating sticks outside might be a reverse causality but may also result in mechanical damage in the gastrointestinal tract and possible perforations by the sticks’ sharp edges, which may cause inflammation. While symptoms have been noted in puppies and young dogs, CE is usually diagnosed later, in middle-aged dogs94,95. The difference between the onset of owner-reported symptoms in our data and the general reported age of diagnosis is likely explained by the progression of the disease from subtle to more severe clinical signs requiring veterinary care, and the time it takes to perform several tests for diagnosis (e.g. exclusion of other causes and histopathology of gut biopsies)6.
The current study has several strengths. The study data were obtained from a partially validated questionnaire which provides reasonable and trusted data. The study took reverse causality into account by excluding the dogs based on age criteria from analyses. A wide range of food items were covered in the questionnaire and the diet style was inferred from several food variables using multivariate methods. Single food items within the diet style and other edible items were also studied. Additionally, two age periods (puppyhood and adolescence) were included.
The present study has some limitations. The study design is an owner-reported longitudinal and cross-sectional study which only can suggest causal relationships, not an experimental study on cause-effect. An owner reported FFQ was used, which may have led to recall bias and misclassification of the food items. However, we have validated the owner's answers by resending the questionnaire for them to refill26. We can thus assume that recall bias was substantially reduced. The FFQ is based on frequencies and not quantity and therefore we could not estimate the amount of individual food items of interest. Also, due to the lack of details regarding the ingredients of the food variables, we were not able to examine the nutrient profiles of the diets. Many owners have left part of the detailed feeding frequency questions unanswered instead of giving the answer “never”. However, the sample size retained was of reasonable size. Unfortunately, no information on antibiotic treatments is available in the questionnaire used. Due to the complexity of this study (different diet styles, 47 individual food items and two life stages) we did not include environmental factors as they were discussed earlier in Hemida et al.12.