This experiment was approved by the University of Guelph's Animal Care Committee (AUP #4365) and was carried out in accordance with national and institutional guidelines for the care and use of animals. Thirty client-owned, adult (7.2 ± 3.1 years) dogs of mixed sex (17 females: 16 spayed, one intact; 13 males: 10 neutered, three intact), weight (27.4 ± 14.0 kg) and breed participated in this study (Table 1). All dogs were deemed healthy based on their previous medical history as well as a pre-study physical examination performed by a licensed veterinarian, complete blood count (CBC), and serum biochemistry profile. During the recruitment process, dogs were excluded if they had any skin conditions, received any pro- or anti-inflammatory medications 2-months prior to baseline samples, had abnormalities on their physical examination, CBC, or serum biochemistry, or were younger than 2 years of age. Dogs were housed at their owners' homes for the duration of the study, they followed their usual daily routines. Pet owners were instructed to provide no supplements, medications, antibiotics, antifungals, antiparasitics, or topical creams without notifying the researchers. Prior to week 10, dog #10, consuming FLX, withdrew from the study due to circumstances unrelated to the research trial or treatment diet.

Over a 4-week wash-in period, all dogs were acclimated to a dry extruded commercial kibble (SUMMIT Three Meat Reduced Calorie Recipe, Petcurean, Chilliwack, BC, Canada; Table 2), sunflower oil (SA Kernel-Trade, Kuiv, Ukraine; Table 3), and beef-based treats (Beef Tendersticks, The Crump Group, Brampton, ON, Canada; proximate analysis: metabolizable energy 3039 kcal/kg; crude protein minimum 65%; crude fat minimum 5.1%; crude fiber maximum 4.0%; moisture max 9.56%). Oil was included in the diet at 8.2 grams of oil per 100 grams of total food intake, bringing the total dietary lipid content to 20% on an as-fed basis. Treats were included in the diet up to 2.5 grams per 100 grams total intake, and the remaining proportion of the diet was provided as kibble. During the wash-in period and throughout the study, daily portions of food, oil, and treats were pre-weighed by researchers and provided to the owners in 2-week intervals to be offered to dogs daily at a frequency determined by the owner. To avoid the occurrence of lipid peroxidation, owners were instructed to mix the oil with the food immediately before feeding. Any leftover kibble, oil, and/or treats were returned to researchers and subsequently weighed and recorded. Dogs were initially fed to meet their estimated maintenance energy requirements (110 kcal ME × kg BW^0.75), and BW was recorded every 2 weeks starting at baseline. Each dog's food allotment was then adjusted accordingly to maintain baseline BW throughout the study. No abnormal observations were reported by owners throughout the 16-week study period in terms of diet tolerance (i.e., vomiting, stool quality, halitosis, etc.).

This study was conducted using a randomized complete block design (RCBD) with repeated measures. Following the 4-week wash-in period, dogs were blocked by breed, age, and BW and groups were randomly assigned to one of 3 treatment oils: camelina oil (CAM) (n = 10; eight females; two males), flaxseed oil (FLX) (n = 10; five females; five males), or canola oil (OLA) (n = 10; four females; six males). The sunflower oil used during the wash-in was replaced with either CAM, FLX, or OLA, and feeding continued as described for 16 weeks. Both OLA and FLX were chosen as control groups for this study as they are commonly used to formulate canine diets and provide a source of n-3 FAs.

Dogs were fasted for a minimum of 10 h overnight and blood samples were collected via cephalic venipuncture using a syringe (Becton, Dickinson and Company, Franklin Lakes, NJ, USA). Of the collected blood, 5 mL was put into a serum vacutainer (Becton, Dickinson and Company, Franklin Lakes, NJ, USA). Blood was allowed to clot and was centrifuged at 7,200 × g for 15 min using an accuSpin Micro 17 centrifuge (Thermo Fisher Scientific, Waltham, MA, USA). Then, the serum aliquots were frozen at −80°C until later analysis.

Serum samples were analyzed for prostaglandin E₂ (PGE₂) (Canine Prostaglandin E₂ ELISA Kit MBS013017, MyBioSource, Vancouver, BC) and junction plakoglobin (JUP) (Canine Junction Plakoglobin ELISA Kit MBS104997, MyBioSource, Vancouver, BC) using commercially available ELISA (Enzyme-linked immunosorbent assay) kits. Samples were run in duplicate according to the manufacturer's instructions. Serum glycosaminoglycan (GAG) (dimethyl methylene blue) and nitric oxide (NO) (Griess Reaction; Molecular Probes, Eugene, OR) concentrations were determined using spectrophotometric assays (26, 27). Serum NO and GAG samples were analyzed as previously described by MacNicol et al. (28).

Skin barrier function and integrity were assessed by measuring TEWL, which is defined as the amount of water that passively evaporates through skin to the external environment due to a water vapor pressure gradient on both sides of the skin barrier and is commonly used to characterize skin barrier function and integrity (29, 30). On weeks 0, 2, 4, 10, and 16, TEWL was measured using a VapoMeter^® SWL-3 (Delfin Technologies Ltd, Kuopio, Finland), according to the manufacturer's instructions. Since privately-owned dogs were used, it was not feasible to shave multiple patches for TEWL measurements, and as a result, researchers chose three body sites with little hair to measure TEWL, including: the right paw pad, right pinna, and right inner thigh. Ten measurements were taken per body site and the average was used for analyses. Once the averages were calculated, any values above or below the average by 50 g/m²/h or more were considered outliers and removed. All dogs were brought to the University of Guelph by their owners on collection days to ensure environmental conditions during collections remained consistent. All measurements were carried out by a single operator, in the same order of body sites, and in a climate-controlled room to maintain consistency between samples and to avoid variation in VapoMeter^® readings due to temperature and humidity fluctuations (29). Room conditions were stable at 22–23°C ambient temperature and 44–50% ambient relative humidity. The evaporation rate value is calculated in grams of water per square meter per hour (g/m²/h). All dogs were behaviorally acclimated to the use of the VapoMeter^®, the researchers involved in sample collection, and the collection room, prior to the first sample day to minimize stress, thereby reducing variation in measurements. If dogs were wet due to weather upon arrival they were dried with a towel, to reduce variation further.

Two researchers blinded to treatment were trained to perform a subjective coat assessment on weeks 0, 2, 4, 10, and 16 using a 5-point Likert scale (under Supplementary material). A Likert scale was used to measure the softness, shedding, dander, shine, spring, softness uniformity, color, color uniformity, and follicle density of the coat. Follicle density was assessed on the center of the back of the dogs by scoring the thickness/amount of hair coming from individual follicles. To increase consistency among dogs given different management practices in each household, all dogs were bathed 2 weeks prior to each assessment and owners were instructed to keep dogs dry and to not brush or groom them during this period.

Data are presented as mean ± SD unless otherwise stated. All statistical analyses were performed using the PROC GLIMMIX of SAS Studio^® software (v.9.4., SAS Institute Inc., Cary, NC, USA). Dog was the experimental unit, and treatment, TEWL site, and sex, and age were treated as fixed effects (age and sex data not presented). Week was treated as a repeated measure. An analysis of variance (ANOVA) was performed to assess the effects of treatment on inflammatory and oxidative marker concentrations, TEWL, and coat scores. When the fixed effects were significant, the means were separated using Tukey–Kramer adjustments. Significance was declared at a P ≤ 0.05. Trends were declared at P ≤ 0.10.

Of the 4,440 individual TEWL measurements collected throughout the study period, 18 were considered outliers and removed [D = Dog, W = Week; Paw pad: D6W2(CAM), D8W16(FLX)(2 values), D9W16(FLX), D17W4(CAM), D18W2(FLX), D18W4(FLX)(2 values), D23W10(CAM), D23W16(CAM); Inner ear: D5W4(OLA), D5W10(OLA), D12W10(OLA); Inner leg: D6W2(CAM), D6W10(CAM), D12W0(OLA), D16W0(FLX), D29W0(FLX)]. These outliers could often be attributed to changes in the environment, leading to signs of stress or excitement in the dogs (i.e., researchers entering and leaving the room, noises occurring outside of the sample room, and in the case of some outliers these samples were taken near the end of the collection period and the dogs would become impatient, no longer wanting to remain in the same spot for samples).

There were no differences among treatments (P = 0.726), across weeks (P = 0.738), or for treatment by week interactions (P = 0.996). Additionally, there were no differences for site by week (P = 0.378), or sex (P = 0.274) (Table 5). However, there were differences observed among sites (P < 0.0001), in that TEWL values for the paw pad were greater than those of the pinna or inner thigh. Additionally, there was a trend observed in age (P = 0.072), in that senior dogs (11–14 years; n = 3) tended to have lower mean TEWL values compared to young (2–4 years; n = 7), young adult (5–7 years; n = 9), and adult dogs (8–10 years; n = 9).

In the current study, concentrations of GAG tended to be higher in males compared to females. Studies in humans by (1) Larking (31) and (2) Claassen and Werner (32) found that, similar to the present study, females have lower concentrations of GAG. Claassen and Werner analyzed GAG in thyroid cartilage while Larking measured GAG excretion in the tissue. Since GAG is a marker of cartilage turnover, Claassen and Werner attribute their findings to greater cartilage turnover in males, while Larking accredits their findings to the males in their study having a greater mean height (31, 32). It is possible that the female dogs in the present experiment had a smaller average height and lower cartilage mineralization than the males, which contributed to the lower concentration of circulating GAGs observed. However, height and cartilage mineralization were not measured in the present study. Furthermore, the observation made in our study was only a tendency; this, combined with the dearth of work carried out in dogs and lack of equal distribution of male/female, intact/neutered/spayed dogs in the current study make it difficult to form any cogent conclusions. Future research should investigate this relationship further using a dog model.

No significant changes were observed in PGE₂, JUP, GAG, or NO concentrations over the 16-week study period. It is possible that the stability of these concentrations across time and among treatments is attributed to the lack of exercise or immune challenge experienced by the dogs on the current study. It is well-established that both exercise and immune challenges result in a wide range of physiological and biochemical adaptations, the magnitude of which is directly related to the intensity and duration of the exercise or immune challenge encountered (33–36). This wide range of physiological and biochemical adaptations include changes in inflammatory and oxidative biomarker concentrations (28, 33).

Dogs and horses both experience increased PGE₂ concentrations following exercise. In horses, NO and GAG concentrations increase following exercise and compared to baseline, but no change was observed in dogs (28, 33). Pearson et al. attribute these results, similar to previous findings, to variations in NO production depending on exercise intensity, suggesting that it is possible that the lack of changes observed in NO concentration in the current study is due to the low intensity of the exercise experienced by the dogs (33). Markers like PGE₂, NO, GAG, and JUP are often upregulated during times of immune challenge/disease (37–40). A myriad of studies completed in humans suggest no effects of n-3 PUFA supplementation on inflammatory or immune markers in healthy individuals (41–43). As an example, Pot et al. found that supplementing fish oil and sunflower oil to healthy individuals had no effect on chemokine, cytokine, or cell adhesion molecule concentration compared to baseline (41). Healthy individuals, similar to the canine subjects of our study, generally have low levels of circulating inflammatory markers. Thus, the chance that low levels of inflammation are reduced even further by an intervention with oil is very small and difficult to measure. The dogs of the present study were healthy upon recruitment and on every sample period based on a veterinary examination, as well as CBC and biochemistry analysis, indicating a lack of immune response that would elicit an inflammatory response. Additionally, the dogs did not participate in any intense exercise prior to or on sample days, and thus had no known reason to elicit any exercise stress induced response impacting markers of inflammatory or oxidative stress. For safety and animal care purposes, no procedures with the potential to cause harm to the animals, like an inflammatory or immune challenge, can be carried out in client-owned dogs. Additionally, the objective of the present study was to determine how these three oils compare to one another in terms of their effects on these biomarkers to gauge their use in dog food formulations for typical pets, not to evaluate their performance following an exercise or immune challenge. Future studies should compare the effects of these three oils and their performance following exercise and immune challenge.

In the present study, mean TEWL values were significantly greater when measured on the paw pad compared to the inner leg and inner ear. This is likely the result of the tubular, unbranched eccrine glands that open directly onto the skin of the paw pads and noses of canines. These glands allow sweat to be released from these areas, contributing to the water-loss detected by the VapoMeter, and thereby likely contributing to greater TEWL values compared to the inner leg and pinna (44). Additionally, TEWL values were found to be lower in senior dogs compared to young, young adult, and adult dogs. Similar findings have been observed in other canine and human studies and although the exact mechanism behind these observations is unclear, there are various theories (45, 46). The thickness of the stratum corneum and flattening of corneocytes increases with age, while natural moisturizing factors, stratum corneum hydration, and epidermal lipid synthesis are reduced (47–53). Additionally, the density of dermal capillaries decreases with age, which may lower skin temperature and in turn decrease water diffusion (51, 54). All of these findings provide examples of mechanisms that increase the path length and resistance of a water molecule and subsequently contribute to lower TEWL in older individuals, and in agreement with the present study.

Spring and follicle density increased significantly from baseline. This is likely due, at least in part, to the growth of winter coats as the study began at the end of summer and went into the winter (September–January). Dogs have a light summer undercoat that is shed before a thick winter undercoat grows in, which could explain the increase in spring and follicle density. This further supports the observation of the present study in that shedding was greater in all dogs at the beginning of the study at weeks 0 and 2, compared to weeks 10 and 16.

Softness, shine, and color of the dogs' coats increased from baseline. This is likely a result of the dogs consuming an increased amount of n-3 FAs following baseline, which can be further metabolized into EPA and DHA, though with limited efficiency. Supplementation of fish oil, a rich source of EPA and DHA, was found to improve skin and hair coat quality in dogs from baseline based on a clinical score, with maximal improvement occurring after 8 weeks (55). The positive effects on skin and coat health are thought to be due to an increase in EPA and DHA in the erythrocyte membrane, along with increased total lipids in the hair shaft (55). The same study observed that following supplement withdrawal, skin and coat health clinical scores remained the same for 1 month and began to deteriorate following the second month (55). Although we did not take measurements on week 8, we did take measurements on week 10, and this is where we saw the largest improvement (i.e., softness, shedding, shine, spring, and color). This is most likely due to the increase in ALA, which is the parent compound of EPA and DHA, the dogs received from their treatment oil (CAM 1:1.8, FLX 1:4.19, OLA 1:0.59) in comparison to the wash-in sunflower oil (1:0). It is important to note that our study had no negative control group, since the absence of an oil supplement would alter all macronutrient intakes and our aim was to compare to existing approved oil supplements. As a result it cannot be ruled out that the observed changes in coat quality may be a result of the placebo effect. Future studies should consider employing a control group fed no oil supplement to rule out the possibility of the placebo effect impacting observations.

All dogs in the current study were considered healthy, with no known dermatological conditions or skin disorders. The coats of these dogs were in relatively good condition at baseline, and future research should investigate these oil supplements and their effects on skin and coat health in dogs with poor skin and coat quality as a result of conditions like atopic dermatitis. It is important to note that ectoparasites, particularly fleas in dogs, can negatively impact skin and coat health (56). In this study, although complete blood count and biochemistry values were assessed, and physical examinations were performed by a licensed veterinarian prior to study recruitment and throughout the entire trial, diagnostic and preventive control in terms of ectoparasites was not considered, and this is a limitation of this study. Authors recommend future studies consider using more specific techniques as inclusion criteria when recruiting participants in order to ensure the absence and prevention of parasites and their potential impact on skin and coat health.