Speed of eating may predispose to excessive food intake in domesticated dogs and, therefore, strategies that prolong the time spent feeding might help to prevent the development of obesity. The aim of the current study was to determine the effect of three different puzzle feeders on feeding behaviour in healthy pet dogs.

Nine healthy pet dogs participated in the study; a range of breeds, ages and genders, were represented, and all were in ideal body condition. Three commercially-available puzzle feeders (PF) (PF1: Original Kong® toy, Kong Company; PF2: Buster® Cube, Kruuse UK Ltd.; PF3: Dog Pyramid^TM, Zoo Active Products) were tested against a control strategy (the dog's own feeding bowl). A standard amount (5% of their usual daily intake) of their usual dry kibbled food was offered with each intervention, on two separate occasions, in a random order determined by Latin-square design. The time taken to empty the puzzle feeder completely was measured, whilst other parameters (e.g., ease of use for dog and owner, aggression, active behaviour and frustration) were each subjectively scored on a 5-point scale. A further study tested the effect of a 30-minute training session on performance of the puzzle feeders in 3 of the dogs.

PF1 increased the time of food consumption by almost twenty times compared with the control intervention (P = 0.0039). Most of the dogs struggled with PF2 and PF3, with only two dogs emptying the feeder completely; the remaining 7 dogs walked away prior to completion and, therefore, median time to completion could not be calculated for either PF2 or PF3. PF1 was significantly easier to use than both PF2 and PF3 (P < 0.001), with no difference between PF2 and PF3 (P = 0.0628). Further, PF1 stimulated greater active behaviour during use (P = 0.002) than both PF2 and PF3, again with no difference between PF2 and PF3. However, there was no difference in aggression or frustration amongst feeders (P > 0.2). After a 30-minute training session, one dog was able to empty both PF2 and PF3 completely, but this training did not affect performance in the other two dogs.

In conclusion, the current study has shown that some puzzle feeders can successfully slow food intake, although design influences ease of use. Long training periods may be required for designs that are inherently less intuitive for dogs to use.