The experiment was carried out using a total of 48 Limousine young growing fattening animals (24 bulls and 24 heifers), aged 250 ± 20 days at the time of the first sampling. The initial body weight (BW) was on average 350 ± 15 kg for bulls and 280 ± 10 kg for heifers. This breed was chosen because it is the most diffused in Sicily for breed production. All animals were housed in the same pen of a semi-open straw-bedded barn located in Santa Croce Camerina (Ragusa, Italy, 36°49′38″N 14°31′26″E). The farm is located within the Hyblean Mediterranean area with an unequally distributed rainfall throughout the year. Annual mean rainfall is about 700 mm. The annual average temperature is 19.58°C (with an average of maximum and minimum temperature of 21 and 17°C, respectively).

Animals were raised according to an approved UE disciplinary called “QS Sicilia,” which allows the inclusion of olive cake in feed for beef cattle not < 10% of the diet, as a strategy for the recovery of agro-industrial by-products.

After a 2-week adaptation period, animals were randomly allocated to one of three dietary treatments according to their BW: inclusion of olive cake at 0% (Control group: CTR), 10% (Low-olive cake group: L-OC), or 15% (High-olive cake group: H-OC) in the concentrate. The compositions of the concentrates and of the OC used for the study are reported in Table 1. All the animals were allowed ad libitum access to straw and water. After adaptation, animals received their concentrate (equal to 2% of their body weight, on average) twice daily (7.00 a.m. and 3.00 p.m.). During the whole experimental period (147 days), none of the animals had health problems. When animals reached the endpoint, they were transported to a licensed slaughterhouse located 20 min (19.2 km) away from the farm. Animals fasted for 12 h before slaughter, which was performed according to EU Regulations. The interval between the first and the last slaughter was 18 days.

Animals of each pen, before feed distribution, were individually weighed at the beginning of the trial, after 56 days, and at the end of the trial (147 days) using Brecknell PS-2000 Veterinary Floor Scale (capacity: 1,000 kg, readability: 0.5 kg). Individual average daily gain (ADG) was also calculated. At the end of the trial, animals were slaughtered, and after slaughter the hot carcass weight (CW) was recorded, and the dressing percentage was calculated. pH was determined 1 and 24 h after slaughter using a portable pH meter (Hanna Instruments, Woonsocket, RI, USA) and combined glass electrode, inserted ~5 cm into the longissimus thoracis muscle.

The blood samples were collected at the beginning (day 0, before administrating treated concentrates) and on days 56 and 147 of the trial, in the morning before dispensing the concentrate. Blood was collected from the tail vein by trained veterinarians through venipuncture into 10-mL tubes containing clot activator and separating gel (Terumo Corporation, Tokyo, Japan). Blood samples were centrifuged for 10 min at 2,000 × g; the supernatant serum was collected and stored at −20°C until analyses.

Biochemical lipids (TG, TCHO), liver (TBIL, AST, ALT, LDH), renal (Urea), and muscular (CK), parameters were analyzed by enzymatic colorimetric method, using an automated spectrophotometry (Biotechnical Instrument BP 3500) and reagents of the same commercial house.

Glucose concentration was measured by automated spectrophotometry (Biotecnica Instruments BT 3500) using the colorimetric enzymatic method GOD-POD.

Insulin concentration was analyzed using the Immulite® two-site chemiluminescent immunometric assay (Maglium 800), with a sensitivity of 0.5 μIU/mL and intra- and inter-assay coefficients of variation equal to 1.56 and 4.07% at insulin concentrations of 16.54 and 45.804 μIU/mL.

Statistical analyses were performed with JMP®, Version 16 (SAS Institute Inc., Cary, NC). Appropriate descriptive statistics was generated for all the variables. The correlation between all the parameters was expressed using Pearson's coefficient (r).

The blood parameters of interest (Y) were modelized as follows:

Where m is the mean, D is the diet (CTR, L-OC, or H-OC), T is the time of sampling (day 56 or day 147 of the trial), S is the sex of the animal (male or female), DT is the interaction between the diet and the time of sampling, DS is the interaction between the diet and the sex, T0 is the covariate representing the parameter level at the beginning of the trial (before administering the supplemented concentrated), and e is the random residual. With regard to the performance parameters, only the diet (D), the sex (S), and their interaction (DS) were included as factors. A logarithmic transformation was applied when necessary. When factors or interactions resulted significant, the Tukey-Kramer post-hoc test was used to identify the significantly different levels.

Differences were considered to be statistically significant when P < 0.05.

In the present study, growth of animals was characterized by an increase in triglyceridemia and total cholesterolemia. The increase in TCHO concentration can be explained by the marked growth of animals during the finishing phase and the consequent intensification of the anabolic activity of fat and liver metabolism (Van Soest, 1994). Moreover, several studies observed that a high level of TCHO during the early stages of growth is also related to the precocity of puberty attainment (Anderson et al., 2015; Rodríguez-Sánchez et al., 2015) and that TCHO concentrations increase with advancing age in beef cattle, according to the start of puberty and related sexual maturity. Concerning the effect of the diet, we also observed that the increase over time of TCHO was to a lesser extent in the H-OC group only. Accordingly, previous studies investigating olive by-products in sheep diets found no variation or a decrease in TG and TCHO concentrations (El-Tarabany et al., 2018; Awawdeh et al., 2020; Alkhtib et al., 2021). This may be due to the OC's high content in unsaturated fatty acids and phytosterols (Viveros et al., 2009; Cedó et al., 2019).

When considering the effect of sex, heifers showed significant higher TCHO and TG concentrations than bulls. Many aspects of sex differences in physiology arise due to the mechanisms of the sex hormones, including TG and TCHO biology. Estrogen has pleiotropic effects on many tissues and pathways that govern lipid and lipoprotein metabolism and also affects cholesterol efflux capacity (Matthews et al., 2000; Zhu et al., 2018). Heifers may potentially have the capacity to undergo different metabolic adjustments according to their sexual maturity, compared to bulls, as confirmed by the significant effect of age at first service on heifers' plasma concentrations of TG and TCHO observed in cattle and buffalos (Talavera et al., 1985; Wehrman et al., 1991; Ryan et al., 1992; Campanile et al., 2010; Hussein and Abdel-Raheem, 2013). This might explain why we could observe a slight difference in TG concentration due to the diet in bulls but not in heifers.

We observed that TBIL concentration increased from 56 to 147 d and was positively related to the OC inclusion. However, it is important to underline that all the measurements remained within the physiological range.

There are few studies reporting the effect of the diet on this parameter, and no study concerns the integration of olive by-products. It is known that, in cattle, a negative energy balance affects fat mobilization and liver function, leading to an increase in TBIL serum concentrations (Mayasari et al., 2019; Marcato et al., 2021; Hisadomi et al., 2022). It has been observed that unsaturated fatty acids, largely present in the OC, reduce bilirubin conjugation, thus increasing its serum concentration (Hargreaves, 1973; Oliveira et al., 2021). In fact, a similar effect was also observed in new-born calves fed with banana extracts (Keivani Rad et al., 2021).

The observed significant increase of urea concentrations from 56 to 147 d is consistent with data previously recorded in both beef and dairy cattle (Swali et al., 2008; Brickell et al., 2009; Gonano et al., 2014). A difference in urea concentration was observed in bulls fed with L-OC and H-OC diet, which showed the highest and lowest values, respectively. Instead, heifers showed similar concentrations regardless the OC inclusion. Urea is one of the few blood parameters that was often evaluated in studies regarding the integration of olive by-products in livestock diet, with most of them reporting a decrease in uremia, which supports our results (Yáñez-Ruiz and Molina-Alcaide, 2007; El-Tarabany et al., 2018; Awawdeh et al., 2020; Alkhtib et al., 2021). This effect was associated to a possible reduction in nitrogen intake or protein rumen degradability due to the bind between tannins and dietary proteins (Yáñez-Ruiz and Molina-Alcaide, 2007; Correddu et al., 2020).

With regard to hepatic enzymes, we observed that ALT was higher in control animals, but the differences among the groups disappeared when we considered the last sampling period. Accordingly, animals of several species, both ruminant (El-Tarabany et al., 2018; Lipińska and Józwik, 2018; Awawdeh et al., 2020; Alkhtib et al., 2021) and monogastric (Paiva-Martins et al., 2014; Parsaei et al., 2014), showed similar or slightly lower ALT activity when fed with olive or chokeberry by-products. It can be hypothesized that this enzyme decreases due to the effect of antioxidants and polyphenols on the liver.

Interestingly, we also observed a similar effect of sex-diet interaction on all the hepatic enzymes. Particularly, lower activity of all of them were measured with the 10% integration (L-OC), in bulls only. Most of the studies on olive by-products administration showed no differences in these parameters (Paiva-Martins et al., 2014; El-Tarabany et al., 2018; Awawdeh et al., 2020; Alkhtib et al., 2021), but the extent of the integration might be important; in fact, in broilers, a small quantity of olive cake has been associated to a decreased AST, whereas a great quantity to increased AST activity (Parsaei et al., 2014). Moreover, a reduction of AST and LDH activities was also recorded in lambs fed with chokeberry by-products (Lipińska and Józwik, 2018). It might be speculated that these alterations depend on a lower diet nitrogen content, which can influence the hepatic metabolism decreasing AST and ALT activities (Puppel and Kuczyńska, 2016).

As well as AST, CK activity was higher in CTR animals than L-OC and H-OC. The CK is an indicator of muscle protein turnover and is also associated with energy utilization (Baird et al., 2012). Its activity can vary according to the time of the day, age, growth rate, and pregnancy (Gonano et al., 2014); nevertheless, in the present study it is possible to exclude the effect of these variables on the changes of CK activity, because animals were homogenous for age, growth rate, time, and physiological conditions.

In addition, elevated CK activity due to the muscle damage, which can be caused by the physical activity-related oxidative stress, can be prevented or reduced by antioxidant supplementation (Ostojic et al., 2008; Wang et al., 2008; Gupta et al., 2009; Marius-Daniel and Stelian Dragomir, 2010; Bentley et al., 2012). For example, red fruit oil, a natural source of antioxidant, showed to decrease significantly the CK activity during exercise when supplemented to mice (Sinaga and Purba, 2018). On this evidence, it is possible to presume that the antioxidants present in the OC contributed to decrease the activity of CK in both L-OC and H-OC. What is more, diets supplemented with olive or chokeberry by-products led to no variation or a decrease of creatin kinase activity, similarly to our study (Lipińska and Józwik, 2018; Awawdeh et al., 2020).

Insulin is known to accelerate anabolic processes, such as the synthesis of muscular protein contributing to muscle development, and the synthesis and deposition of fat along developmental processes. Insulin is essential for the regulation of glucose and lipid metabolism and also affects growth through its promotion of the uptake of nutrients into body tissues as recorded by Martin et al. (Martin et al., 1984) and as shown by the existence of significant correlations between insulin and glucose, TCHO, TG, AST, and TBIL. On the other hand, glucose is the favored substrate for lipogenic adipocytes in muscle, whereas acetate is preferred in subcutaneous adipocytes and as precursor of de novo fatty acid synthesis (Smith and Crouse, 1984; Gilbert et al., 2003; Ladeira et al., 2016; Bionaz et al., 2020).

In the present study, insulin was higher in bulls and increased from 56 to 147 d, whereas the opposite was true for glucose, in accordance with the literature (Roy et al., 1983; Gray et al., 1986; Beeby et al., 1988; Plouzek and Trenkle, 1991; Shingu et al., 2001). Bulls, in fact, show higher concentrations of insulin-like growth factor 1, growth hormone, and insulin compared with cows, probably due to the effect of sex-dependent steroid, and this might partially explain their increased growth rate (Ronge and Blum, 1989; Plouzek and Trenkle, 1991; Sirotkin et al., 2002). Fat accumulation in different depots is also sexually dimorphic: in humans, men accumulate more visceral fat, whereas women accumulate more subcutaneous fat and have a higher overall percentage of body fat. There is also evidence indicating that insulin sensitivity differs between males and females (Sirotkin et al., 2002; Macotela et al., 2009). In mice, it was observed that the intra-abdominal depot is regulated by physiological levels of sex steroid and that female mice are more insulin-sensitive that males and their adipocyte show higher expression of glucose and lipid metabolism genes (Mittendorfer, 2005). However, how insulin action differs between males and females and how these differences account for a sex-specific regulation of adipose tissue development and function are not completely clear.

In the present study, we observed a similar insulin concentration in heifers and bulls of the L-OC group, even if diet or its interaction had no effect on glycemia. It should be noticed that, as far as we know, this is the first study analyzing the effect of an olive by-product inclusion on insulin in livestock, but on the other hand it has been demonstrated the phenols contained in olive leaves have a hypoglycemic effect in several species (Parsaei et al., 2014; Ait-Kaki et al., 2018; Lipińska and Józwik, 2018; Alkhtib et al., 2021). Moreover, it is known that a diet rich in carbohydrates tends to increase insulin (Mori et al., 2007) and that glucose kinetics is influenced by the intake of nitrogen and nutrients as well as the fasting length (Brickell et al., 2009; Zanton and Heinrichs, 2017); therefore, we cannot exclude that other mechanisms can be involved in the observed insulin alteration.