The effects of a plant-based and a plant- and marine-based n-3 oil supplement on behavioural reactivity, heart rate variability, and plasma fatty acid profile in young healthy horses

Abstract

Behavioural reactivity in horses poses a welfare and safety risk to both the horse and the handler, however, beneficial effects have been observed when dietary fat is increased in replacement of sugar. Supplementation with the fatty acids (FA) eicosapentaenoic (EPA) and docosahexaenoic acid (DHA) appear to improve negative behaviours in rodents and humans, but the effect of α-linolenic acid (ALA), EPA, and DHA, specifically, on reactivity in horses is unknown. The objective of this study was to evaluate the effects of camelina oil (CAM; ALA-enriched) and a mix of camelina and algal oil (ALG; ALA-, EPA-, and DHA-enriched) both fed at a dose of 0.37 g oil/kg body weight on plasma FA, behaviour, and heart rate variability (HRV) in young horses compared to a negative control (CON). Thirty-four client-owned horses aged 7 months – 6 years were enrolled. Horses were assigned to either CAM, ALG, or CON and underwent a novel object test (NOT) before and after a 6-week supplementation period. Prior to each NOT, blood was collected for evaluation of plasma FA profile (n = 28). During the NOT, behaviour was recorded using a pre-determined ethogram and assessed in BORIS software by two raters (n = 29). Electrocardiogram (ECG) data was collected at baseline, during the NOT, and after the NOT (recovery). The ECG data was analyzed in Kubios software for determination of heart rate (HR) and several HRV parameters (n = 24). The treatment oils were treated as fixed effects, baseline measurements as covariates, and location as a random effect. Plasma DHA (P < 0.01) was greater and n-6:n-3 ratio (P < 0.01) was reduced in ALG than in CAM and CON, while ALA and EPA were similar among treatments (P > 0.05). When treatments were pooled, the maximum HR (P < 0.01) and the low frequency to high frequency ratio HRV parameter (P < 0.01) were greater during the NOT compared to baseline and recovery. Bucking (P = 0.03) and backing (P = 0.02) behaviours were reduced in the CAM group compared to the CON group, but neither group differed from ALG. All other behaviours, HR, and HRV parameters were similar among treatments (P > 0.05). Our results suggest that the NOT was successful in creating acute stress, however, feeding either CAM or ALG at this dose did not reduce reactivity in this cohort of horses. Further research is needed to understand the effects of specific FA, if any, on behaviour and HRV in more specific populations of horses and specifically those deemed highly reactive.