Dietary format affects selenium absorption and bioactivity in dogs

Selenium (Se) is an essential trace element with antioxidant properties that protects the body against oxidative stress. A dietary safety margin is needed to avoid loss of this protection at low concentrations and development of Se toxicity at high concentrations. The current recommended allowance for Se in dog foods is not format specific, but it is known that canned and kibble diets differ in the raw materials that are used and the way the diets are processed. Raw materials vary greatly in their Se concentration and availability. However, little is known about which dietary factors affect Se absorption, i.e. the fraction of dietary Se that reaches the systemic circulation, and Se bioactivity, i.e. the amount of Se that can be used for the incorporation into enzymes, in dog foods. Our previous work has identified diet format and apparent crude protein (CP) digestibility as factors that affect in vitro Se accessibility. Based on the in vitro results, it was hypothesized that canned diets have a lower Se absorption than kibble diets and that CP concentration is positively associated with Se absorption in kibble diets and negatively in canned diets in vivo. Twenty-four Labrador retrievers were divided into four treatment groups of 6 dogs. The study consisted of four feeding periods, each starting with six days to wean the dogs to their new diets, followed by periods of between 29-43 days for every diet. Eight commercially available diets were selected that varied in format (canned or kibble) and CP concentration. For every format, diets with an intended concentration of 9.6, 14.3, 19.1 and 23.9 g CP/MJ ME (40, 60, 80 and 100 g CP/1000 kcal ME, resp.) were selected. Diets with different CP concentrations were selected, hence leading to a range in digestible CP intakes. During the trial each dog was fed four of the eight experimental diets in a randomised incomplete cross-over design. During the final 6 days of each feeding period, titanium dioxide (TiO2) was added as an inert digestibility marker. Blood, urine and faecal samples were collected at the end of every feeding period. Blood was analysed for whole blood glutathione peroxidase (GPx), serum Se, and general haematology and biochemistry parameters in plasma. Urine samples were analysed for total Se and creatinine (CT) and faeces for TiO2, dry matter (DM), ash, CP (N×6.25) and Se. Apparent Se absorption was calculated as the percentage of Se intake that was not found in the faeces. Data were analysed using linear mixed effects models to investigate the effects of actual CP intake, format and their interaction. Apparent Se absorption (as % of Se intake) was lower in canned compared to kibble diets (p<0.001), which is in accordance with our in vitro Se accessibility findings. Unlike in vitro, no effect of CP intake on apparent Se absorption was found (p=0.753). Urinary Se excretion relative to Se intake was lower in canned than kibble diets (p<0.001) and tended to decrease with increasing CP intake (p=0.059). Even when urinary Se:CT ratio was corrected for the absolute amount of absorbed Se, the excretion was lower in canned diets (p=0.001) and decreased with increasing CP intake (p=0.017). The potential bioactivity was higher in canned than in kibble diets as measured by GPx (p<0.001) and this also decreased with increasing CP intake (p