Materials and Methods

Study design—All dogs were cared for in accordance with internal institutional animal care and use committee protocols. Forty-three weaned 2-month-old large-breed puppies (19 Labrador Retrievers [11 males and 8 females] and 24 Golden Retrievers [13 males and 11 females]) were used. Puppies were reared individually in homes of volunteers^a for the duration of the study. Puppies were allowed enrichment toys and were integrated into the daily lives of their caretakers. The dogs were randomly assigned to 1 of 2 food treatment groups on the basis of litter origin and sex to balance males and females as best as possible without having 1 litter overrepresented in either group. Puppies were fed the assigned food until 18 months of age. Blood samples were obtained at study entry (2 months of age, prior to food assignment) and at 3, 5, 12, and 18 months of age. Dual energy x-ray absorptiometry was performed during general anesthesia at 2, 5, 12, and 18 months of age.

Nutrition—Foods were randomly assigned after evaluation of the 2-month-old dogs, and the caretakers were masked with regard to treatment. Eighteen puppies were randomly assigned to food A^b and 25 puppies to food B.^c Both foods were approved for growth by the Association of American Feed Control Officials Incorporated and consisted of dry kibble and provided a similar caloric density (Appendix). Food was offered in portions in accordance with feeding guides recommended by the manufacturers. No other foods or treats were allowed to be fed for the duration of the study. Food offerings were adjusted to maintain the dog's body condition score close to 3/5. All puppies received standard vaccinations and were spayed or neutered at 6 months of age by their local veterinarian.

Inclusion and removal criteria—Puppies were admitted into the study if they had normal results of physical examination, absence of lameness, and CBC and serum biochemical profile values within reference ranges. The study protocols stated that if puppies were found to have a medical or orthopedic disease during screening, they would be removed from the study and receive treatment appropriate for their condition. Similarly, if puppies failed to consume their food, or if they consumed < 25% of their assigned food for > 4 days or did not gain weight in parallel with their group, they would be removed from the study.

If a dog developed a sustained lameness, it would be examined by a veterinarian and radiography would be performed to determine the cause of the lameness. A dog with either sustained lameness or documented orthopedic disease would be removed from the study and would receive appropriate treatment.

Serum analysis—Blood samples were submitted for CBCs and determination of taurine concentrations. Serum was harvested and stored at −20°C until analysis via biochemical profile and determination of concentrations or activities of vitamin E, fatty acids,¹⁰ glutathione peroxidase, CPII,^d COMP,^e CTXII,^f BAP,^g and osteocalcin.^h Concentrations of 2 hormones, ghrelinⁱ and growth hormone,^j were also measured.

DEXA—A fan-beam type DEXA machine^k that used standard methods was used to examine dogs at 2, 5, 12, and 18 months of age. Standard protocols for body composition based on recommendations by the manufacturer were used.^k The same standard was used to calibrate the machine prior to all measurements.

Statistical analysis—Data were analyzed by use of the general linear models procedure of a software program¹¹ to determine treatment means. Differences were considered significant at P < 0.05.

Results

Dogs—No puppies had disease or lameness at screening, and no puppies failed to consume their food or gain weight. During the course of the study, no dogs had sustained lameness warranting an unscheduled veterinary examination or radiography.

Food proximate analysis—Foods were found to be similar on the basis of acceptable analytic variance as described by the Association of American Feed Control Officials Incorporated, and both were considered appropriate for growth.¹² Food B contained greater amounts of protein, fat, calcium, phosphorus, threonine, methionine, linoleic acid, EPA, DHA, total n-3 and n-6 fatty acids, carnitine, vitamin E, and vitamin C, whereas food A contained more lysine. The quantities of taurine were similar in the 2 foods. The ratio of n-6:n-3 fatty acids in food A was 8:1 and in food B was 2:1.

Serum analysis—Serum biochemical values and CBC values were within reference ranges for both groups at the beginning and throughout the study. Serum concentrations of n-3 fatty acids (A-linolenic acid, EPA, and DHA) and n-6 fatty acids (linoleic acid and arachidonic acid) were tabulated (Table 1). At 2 months of age, prior to dietary intervention, serum concentrations of n-3 fatty acids were not significantly different between groups. Serum concentrations of n-3 fatty acids were significantly greater in puppies fed food B at all subsequent time points.

At the time of screening in 2-month-old puppies, serum concentrations for n-6 fatty acids were significantly greater in puppies assigned to consume food B, compared with those assigned to consume food A. Concentrations of linoleic acid in puppies fed food B were significantly lower than those fed food A at all subsequent time points, and serum concentration of arachidonic acid in puppies fed food B were significantly greater than those fed food A at all subsequent time points.

No significant differences were detected in concentrations or activities of vitamin E, glutathione peroxidase, or taurine in blood in puppies assigned to consume food A versus puppies assigned to consume food B prior to dietary intervention (Table 2). Puppies fed food B had significantly greater serum concentrations of vitamin E, compared with puppies fed food A, at all subsequent time points. Glutathione peroxidase activity was significantly lower in dogs fed food B at 5, 12, and 18 months of age, compared with puppies fed food A. Puppies fed food B had significantly greater concentrations of taurine at 3 and 5 months of age but not at 12 and 18 months of age, compared with puppies fed food A.

No significant difference in growth hormone concentration was detected between groups throughout the study. Puppies fed food B had significantly higher concentrations of ghrelin at 12 and 18 months of age, compared with puppies fed food A.

Concentrations of BAP, CPII, and CTXII were not significantly different between treatment groups at screening and throughout the study (Table 2). Puppies assigned to consume food A had significantly higher concentrations of osteocalcin at 2 months of age, compared with those assigned to consume food B, but the groups were not significantly different throughout the remainder of the study. In both groups, BAP, osteocalcin, and CTXII concentrations decreased as the puppies grew, whereas concentrations of CPII and COMP were similar throughout the study. Concentrations of COMP were not significantly different between groups from 2 through 12 months of age. However, at 18 months of age, puppies fed food B had significantly lower concentrations of COMP, compared with puppies fed food A. The ratio of CPII:CTXII was approximately 2.0 at 2, 3, and 5 months of age in both groups but increased to 7.1 ± 3.5 (mean ± SD) for dogs fed food A and 9.2 ± 4.0 for dogs fed food B at 12 months of age. This difference became significant at 18 months of age, when the ratio was 21.9 ± 17.9 for dogs fed food A and 60.7 ± 76.3 for dogs fed food B.

DEXA and weights—Body weights that were measured versus estimated via DEXA were not significantly different within or between groups throughout the study. At 2 months of age, puppies assigned to consume food B had a significantly higher percentage body fat than those assigned to consume food A. At all subsequent time points, puppies fed food B had significantly less total body fat and higher percentage lean body mass than those fed food A. No significant difference was detected between groups for any of the other variables measured by use of DEXA, including lean body mass, bone mineral content, and bone mineral density, at any time points.

Discussion

Results of the present study indicated that concentrations of BAP, osteocalcin, and CTXII all decreased with age, whereas concentrations of CPII appeared similar at all time points. There did appear to be an effect of nutrition on the ratio of CPII:CTXII but not a significant difference in the means of the individual CPII and CTXII measures. The ratio reflects the ratio of cartilage formation to degradation. This may reflect a positive influence of nutrition on overall metabolism or bone and cartilage metabolism specifically. The histologic and biomechanical properties of the bones and cartilage of these dogs were not examined, so it could not be determined whether the difference in foods altered the composition of these tissues. However, the decrease in COMP at 18 months of age and the greater ratio of CPII to CTXII in dogs consuming food B implied that nutrition may affect cartilage by decreasing basal degradation in favor of formation at maturity. Decreases in COMP and CTXII have been used as prognostic indicators in rheumatoid arthritis and osteoarthritis in humans.^13,14

The addition of n-3 fatty acids appears to increase collagen synthesis and bone modeling-remodeling, which may yield more healthy cartilage and bone in young growing animals entering adulthood. Experiments conducted with growing chicks and dogs have revealed that increasing n-3 fatty acids increase histomorphometric measurements of bone modeling.^15–17 In addition, cell culture studies with n-3 fatty acids and epiphyseal cartilage and chondrocytes from chicks indicate that fatty acids affect cartilage metabolism to increase bone modeling. Other studies^17,18 evaluating the effects of n-6 and n-3 fatty acids on collagen synthesis and PGE₂ production reveal that n-6 fatty acids (linoleic and arachidonic acid) reduce collagen synthesis, but n-3 fatty acids (EPA) appear to stimulate collagen synthesis.

Dogs of large and giant breeds are predisposed to developmental skeletal disease, presumably because of their high growth rate.^3,4 Knowledge of the typical concentrations of bone and cartilage markers obtained from the present study may be used in future studies of their use as markers for developmental disorders.

The differences in n-6 fatty acids at baseline in the puppies reported here were likely related to their dam's food and milk composition, which was not controlled prior to the study; those differences were not likely indicators of inflammation in the 2-month-old dogs. It is not surprising that blood variables reflected the higher concentrations and ratios of n-3:n-6 fatty acids and antioxidants in the 2 foods thereafter.¹⁹ Similarly, because both foods contained similar amounts of energy and were formulated for growth, it is not surprising that consumption of both foods resulted in similar body weights, bone mineral content, and bone mineral density, as measured via DEXA.

Consumption of the 2 foods did appear to affect the absolute fat mass, percentage lean body mass, and percentage body fat in the 2 groups. Puppies fed food B had significantly lower total body fat mass and higher percentage lean body mass than their contemporaries fed food A throughout the duration of the study. This correlated with increased ghrelin concentrations at 12 and 18 months of age in dogs that consumed food B.

Ghrelin is an intestinal peptide involved in growth hormone secretion and energy homeostasis.¹⁸ Fasting ghrelin concentration is negatively correlated with body fat mass and helps regulate food intake in humans.^20,21 In the present study, ghrelin concentration was significantly higher at 12 and 18 months of age in puppies fed food B. This agrees with previous research revealing that animals with lower body fat have higher circulating concentrations of ghrelin.²¹ Additionally, ghrelin directly stimulates bone formation.^22,23 Increased concentrations of ghrelin significantly increase osteoblast-like cell numbers and DNA synthesis in rats²² and increase proliferation and differentiation of rat osteoblasts in cell culture.²³ However, in the present study, a corresponding difference in BAP or osteocalcin in dogs consuming food B was not detected, despite the increased ghrelin concentrations. Therefore, it appeared that ghrelin concentration was more closely correlated with lean body mass (as evaluated via DEXA) than markers of bone turnover.

Loss of articular cartilage, like bone, results from an imbalance between synthesis and degradation over time. C-propeptide of type II collagen is released into the circulation during the formation of mature collagen fibrils and is therefore a marker of cartilage formation,²⁴ whereas CTXII is released during collagen degradation and has been used as a marker of progression of osteoarthritis and may reflect the rate at which articular cartilage is being destroyed.^25–28 Therefore, the ratio of CPII:CTXII may be used to predict the progression of joint damage.²⁹ Indeed, the increased ratio of CPII:CTXII in dogs at 18 months of age in the present study warrants future studies in growing large-breed dogs to examine whether it could be predictive of the onset of osteoarthritis in adulthood.