Arachidonic acid (ARA) is essential in felines because conversion of dietary linoleic acid (LA) to ARA is rate-limited by low Δ6-desaturase. Dietary γ-linolenic acid (GLA) may serve as an ARA precursor by-passing this initial rate-limiting step. This possibility was investigated using twenty-six adult female domestic shorthair cats divided into three groups and fed on complete and balanced diets containing high GLA (GL), high LA (HL) or low LA (LL, control) diets, for 300 d prior to ovariohysterectomy. Plasma was obtained 1–2 d before surgery and uterine, ovarian and associated adipose tissues were reserved for lipid analysis. Fatty acid profiles of the plasma phospholipid (PL) fractions and adipose lipids were performed. In the GL group, plasma and uterine tissue PL were significantly enriched in GLA, di-homo GLA (DGLA) and ARA compared with control. However, ovarian and adipose tissue PL were only enriched in DGLA. Enrichment of uterine tissues with DGLA and ARA probably supplies the essential eicosanoid precursors for reproduction when GLA is fed consistently with an active Δ5-desaturase in uterus. By contrast, this enzyme appears less active in ovary because ARA was not higher compared with control. Earlier reports concluded that ARA was not necessary for fertilisation (an ovarian function), but required for successful pregnancy and reproduction (a uterine function). Adipose tissue DGLA may be a reservoir for ARA synthesis by other tissues upon mobilisation. Dietary GLA may meet feline ARA requirements in the absence of an animal-based preformed source of ARA.

An opportunity to investigate a low-arachidonic acid (AA) feline diet possibly related to elevated peroxide value (PV) during storage on plasma phospholipid (PL) and reproductive tissue fatty acid (FA) profiles presented itself in the present study. Cats (nine animals per group) had been fed one of three dry extruded, complete and balanced diets for 300 d before spaying. The diets contained adequate AA (0·3 g/kg), similar concentration of antioxidants and were stored at ambient temperature, but differed in FA composition. The diets were designated as follows: diet A (high linoleic acid), diet B (high γ-linolenic acid) and diet C (adequate linoleic acid). Diet samples that were obtained the week before spaying revealed an elevated PV of diet A v. diets B and C (135 v. 5·80 and 2·12 meq/kg fat, respectively). Records revealed decreased food consumption of diet A cats beginning at 240 d but without weight loss; thus an opportunity presented to investigate diet PV effects. Total plasma protein and PL-AA concentrations in group A were significantly decreased at 140 and 300 d. Uterine and ovarian tissues collected at surgery revealed modest decrements of AA. Diet A was below minimum standards at 0·015 % (minimum 0·02 %), probably due to oxidation. The time at which diet A became unacceptable may have occurred between 60 and 140 d because plasma PL-AA was within our normal colony range (approximately 4–7 % relative) after 56 d of feeding. High-linoleic acid-containing diets may be more likely to be oxidised requiring additional antioxidants. The findings suggest that reduced plasma protein in combination with plasma AA concentrations may serve as biomarkers of diet peroxidation in cats before feed refusal, weight loss or tissue depletion.