To reduce publication bias and ensure the quality of the meta-analysis, the present study was conducted following PRISMA guidelines (32), as shown in Figure 1. A systematic search for information was conducted using Web of Science, Scopus, PubMed, and ScienceDirect databases to identify previous studies evaluating the effects of dietary supplementation with FLAs on nutrient digestibility, animal performance, carcass characteristics, antioxidant status in blood serum, ruminal fermentation, as well as meat and milk quality in beef (Holstein, Simmental, Angus × Nellore, Jinjiang, Xianan, and native) and dairy cattle (Holstein). The keywords that were used in all the databases were the following: “flavonoids, beef cattle, growth performance, finishing steer, finishing bull, carcass, meat quality, dairy cattle, milk production, milk quality, digestibility, ruminal fermentation, antioxidant status”, and the main representatives of FLAs (33), such as “daidzein, naringin, puerarin, anthocyanin, and quercetin”. In all searches performed, results were restricted to studies published between January 2010 and November 2022. In total, 1,010 scientific publications were identified (Figure 1); however, duplicate publications found in more than one of the databases were excluded. After this, the remaining publications underwent a two-step selection process, as previously reported by other authors (34–36). First, based on the titles and abstracts of each publication, we excluded studies that were not conducted in beef cattle or dairy cows, studies that did not measure any of the variables of interest, in vitro experiments, studies that used animals experimentally infected, as well as simulation and review articles.

Secondly, the articles analyzed had to meet some previously defined inclusion criteria to be included in the final database. In the present meta-analysis, the inclusion criteria used were similar to those previously reported by Dorantes-Iturbide et al. (35) and Orzuna-Orzuna et al. (36): (1) studies with beef cattle or dairy cows housed in confined conditions; (2) data on animal performance, nutrient digestibility, antioxidant status in blood serum, carcass characteristics, ruminal fermentation or quality of the derived products (meat or milk); (3) studies that had control and experimental treatments with similar diets, except for the presence of FLAs in the diets; (4) studies that reported the doses of FLAs used or had sufficient information to estimate the amount of FLAs included in the diets; (5) studies written and published in English and in peer-reviewed scientific journals; and (6) studies that reported the means of the control and FLA-supplemented treatments, the standard error or standard deviation, and the number of replicates.

After applying the inclusion criteria, only 36 peer-reviewed articles were included in the final database (Supplementary Table S1). Likewise, we only extracted quantitative data for response variables that were reported in at least three individual studies (31, 35, 36). Among the response variables included in the final database of this meta-analysis are the following: dry matter intake and nutrient digestibility (neutral detergent fiber, crude protein), daily weight gain, feed conversion ratio, carcass characteristics (carcass yield, backfat thickness), serum concentration of malondialdehyde and antioxidant enzymes (for example, superoxide dismutase), serum immunoglobulins (IgA, IgM, and IgG), rumen parameters (pH, ammonia nitrogen), physicochemical characteristics of the meat (pH, shear force, color), milk production, and milk composition (lactose, protein, and fat content).

Additionally, when available, the following complementary information was obtained from the selected publications: (1) author and year of publication; (2) period of supplementation with FLAs (days); (3) type of FLAs (for example, anthocyanin, daidzein); (4) method of inclusion of the FLAs (extract or naturally present in the diet); (5) amount of concentrate included in the diets (g/kg DM); (6) days in milk from dairy cows; (7) type of cattle (beef cattle or dairy cow); (8) nutritional composition of the diets used; and (9) country where the study was conducted.

Supplementary Table S1 shows the complete list of publications included in the final database of the present meta-analysis. The number of replicates, means, and standard deviations (SD) for the control and experimental treatments (supplemented with FLAs) were extracted from each of these publications. In all the publications in which the SD was not reported, SD was determined using the standard errors of the treatment means (SEM), by using the following equation (37): SD = SEM × √n, where n = number of repetitions.

Meta-analysis and meta-regression, as well as analyzes of subgroups, heterogeneity, and publication bias, were performed using the “metaphor” package (38), which is available in the statistical software R (version 4.1.2, R Core Team, Vienna, Austria). The effects of including FLAs in diets of beef cattle and dairy cows were evaluated using the weighted mean differences (WMD) between treatments supplemented with FLAs (diets with FLAs) and control treatments (diets without FLAs). For this, the means of the treatments were weighted by the inverse of the variance, according to the method for random effects models previously proposed by DerSimonian and Laird (39). In the present meta-analysis, the WMD was used because it allows interpretation of the results obtained in the original units of measurement (40). Additionally, with the PROC MEANS procedure of the statistical software SAS (41), descriptive statistics values were obtained for the continuous covariates level of concentrate in the diet, dose of FLAs, experimental period, and days in milk.

The presence of heterogeneity between studies was identified with the chi-square (Q) test, in which a significance level of p ≤ 0.10 was used since this test has relatively low power (42). Additionally, to quantify the proportion of observed heterogeneity, we used the I² statistic (29). For this test, I² values <25% indicate that the degree of heterogeneity is low, I² values between 25 and 50% indicate moderate heterogeneity, while I² values >50% indicate high and significant heterogeneity (29, 43). On the other hand, to detect the presence of publication bias, the Egger regression asymmetry test (44) was applied, in which a significance level of p ≤ 0.05 was used. When publication bias was detected (p ≤ 0.05 in Egger's test), the “trim and fill” method of Duval and Tweedie (45) was applied to estimate the possible number of missing observations.

Meta-regression analyses were performed on some of the variables evaluated to identify the presence of possible sources of heterogeneity. The criteria considered to apply meta-regression analysis were: (1) presence of significant heterogeneity (i.e., p ≤ 0.10 for Q or I² > 50%); (2) p-value > 0.05 for Egger's test (44); and (3) response variables reported in 10 or more individual studies (46). For all meta-regression analyses, the method of moments of DerSimonian and Laird (39) was used, as it is well-established for estimating between-study variance. Subsequently, for the covariates analyzed that were significant with p ≤ 0.05, the WMD was evaluated through subgroup analysis. A subgroup assessment was not performed when an individual stratum has less than two effect sizes in the meta-analysis (35, 36). The type of FLAs (daidzein, anthocyanin, puerarin, naringin, quercetin, catechin, and blend), the method of supplementation with FLAs (extract or naturally present in an ingredient in the diet), and the type of cattle (beef cattle or dairy cow) were used as categorical covariates. On the other hand, the duration of the experimental period (days), the days in milk of the dairy cows, the content of concentrate in the diet (g/kg of DM), and the doses of FLAs were used as continuous covariates. When any categorical covariate (type of FLAs, type of bovine, and method of supplementation with FLAs) was found to be statistically significant (p ≤ 0.05), subgroup analysis was used to assess WMD (34, 35). Likewise, when the meta-regression was significant (p ≤ 0.05) for the continuous covariates, these were analyzed using the following subgroups: dietary dose of FLAs (≤600 and >600 mg/kg of DM), level of concentrate in the diet (≤400, 401–700, and >700 g/kg of DM), days in milk (≤100 and >100 days) and period of supplementation with FLAs (≤75 and >75 days).

The studies included in this meta-analysis were conducted in eight different countries, mainly in China (44.4%), Spain (16.7%), Brazil (13.9%), and Japan (5.5%). Regarding the animal species (bovine), in 66.7% of the studies, beef cattle were used, and in the remaining studies (33.3%), dairy cows were used (Supplementary Table S1). Supplementary Table S2 shows that the doses of FLAs used were between 12 and 3,104 mg/kg DM. Dairy cows had between 7 and 164 days in milk and the experimental periods ranged between 24 and 168 days (Supplementary Table S2). Supplementary Table S1 shows seven different types of FLAs used in the present meta-analysis. Most of the studies used mixtures of FLAs (36.1%), daidzein (16.7%), anthocyanin (16.7%), and naringin (16.7%). Three other different types of FLAs were used in the remaining studies (13.8%). In addition, 61.1% of the studies used FLAs extracts, and 38.9% used plants or by-products naturally high in FLAs.

Dry matter intake (DMI) increased in response to FLAs supplementation (Table 1). Likewise, dietary supplementation with FLAs increased (p < 0.05) dry matter digestibility (DMD), organic matter digestibility (OMD), crude protein digestibility (CPD), neutral detergent fiber digestibility (NDFD), acid detergent fiber digestibility (ADFD), and ether extract digestibility (EED).

Table 2 shows that daily weight gain (ADG) and backfat thickness (BFT) increased in response to dietary supplementation with FLAs (p < 0.05). In contrast, the dietary inclusion of FLAs decreased the feed conversion ratio (FCR; p = 0.050). However, hot carcass weight (HCW), hot carcass yield (HCY), and Longissimus dorsi muscle area (LDMA) were not affected by FLAs supplementation (p > 0.05; Table 2).

Table 3 shows that dietary supplementation with FLAs increased (p < 0.01) the serum concentration of superoxide dismutase (SOD), catalase (CAT), glutathione peroxidase (GPx), and total antioxidant capacity (TAC). In contrast, a lower (p < 0.001) serum concentration of malondialdehyde (MDA) was observed in animals supplemented with FLAs. On the other hand, the serum concentration of immunoglobulin A (IgA), immunoglobulin G (IgG), and immunoglobulin M (IgM) increased in response to dietary supplementation with FLAs (p < 0.01).

Table 4 shows that the pH and the ruminal concentration of ammonia nitrogen (NH3-N), acetate, and butyrate were not affected by dietary supplementation with FLAs (p > 0.05). However, a higher (p = 0.008) rumen concentration of propionate and a lower (p = 0.023) concentration of total protozoa were observed in response to supplementation with FLAs.

Dietary supplementation with FLAs did not affect (p > 0.05) pH, cooking loss (CL), lightness (L^*), redness (a^*), or meat protein, moisture, and ash content (Table 5). On the other hand, FLAs supplementation decreased (p < 0.05) the shear force (ShF), the malondialdehyde content (MDA), and the yellowness (b^*) of the meat. However, meat's intramuscular fat content (IMF) increased in response to FLAs supplementation (p = 0.029).

Dietary supplementation with FLAs increased (p < 0.01) milk production and milk protein and fat content (Table 6). However, the lactose content in milk was not affected by FLAs supplementation (p > 0.05). In addition, lower milk somatic cell (SCC) counts were observed in response to dietary supplementation with FLAs (p < 0.001).

Tables 1–6 show no publication bias since the Egger regression asymmetry test was not significant (p > 0.05) for any of the variables evaluated. On the other hand, Tables 1–6 show that there was significant (p ≤ 0.10) heterogeneity (Q) for DMI, ADG, FCR, HCY, BFT, SOD, CAT, GPx, MDA in blood serum, IgA, IgG, IgM, rumen pH, NH3-N, acetate, propionate, total protozoa, meat pH, CL, ShF, L^*, protein content, meat IMF and moisture, milk yield, and protein, fat, and SCC content in milk. However, to obtain reliable results, meta-regression analyses are only recommended when the variable of interest is reported in 10 or more studies (46). Consequently, meta-regression analyses were only performed for the following variables: DMI, ADG, rumen pH, acetate, propionate, SOD, milk yield, and milk protein and fat content.

Table 7 shows that the FLAs dose explained (p < 0.05) 23.80% of the observed heterogeneity for milk production. The supplementation period explained (p < 0.05) 5.90, 28.28, 6.87, and 39.61% of the heterogeneity observed for DMI, ADG, rumen propionate concentration, and milk protein content, respectively. On the other hand, the level of concentrate in the diet explained (p < 0.05) 40.10, 37.92, and 10.18% of the heterogeneity observed for ADG, ruminal pH, and milk protein content, respectively. The type of FLAs used explained (p < 0.05) between 20.69 and 56.70% of the observed heterogeneity for DMI, rumen pH, SOD, milk yield, milk protein, and milk fat content. Likewise, the FLAs inclusion method explained (p < 0.05) 15.12 and 40.45% of the observed heterogeneity for SOD and milk production, respectively. Bovine type explained (p < 0.001) 40.60% of the observed heterogeneity for SOD, and days in milk explained (p < 0.001) 41.75% of the observed heterogeneity for milk fat content. There was no significant relationship (p > 0.05) between the covariates used and the ruminal acetate concentration.

DMI increased (WMD = 0.532 kg/d; p = 0.038) when dietary supplementation with FLAs lasted up to 75 days (Figure 2A). However, supplementation with FLAs for more than 75 days did not affect DMI (WMD = 0.015 kg/d; p = 0.805). Higher ADG (WMD = 0.093 kg/d; p < 0.001) was observed when cattle were supplemented with FLAs for periods up to 75 days (Figure 2B). However, when supplementation with FLAs lasted more than 75 days ADG was not affected (WMD = 0.017 kg/d; p = 0.206). Ruminal propionate concentration was increased (WMD = 1.962 mol/100 mol; p = 0.043) in animals supplemented with FLAs for up to 75 days (Figure 2C). However, ruminal propionate concentration was not affected when FLAs were offered for more than 75 days (WMD = 0.306 mol/100 mol; p = 0.486). The protein content in milk increased (p < 0.05) regardless of the period of supplementation with FLAs used (Figure 2D). However, the effect was greater when FLAs supplementation lasted longer than 75 days (WMD = 0.113/100 g) than when it lasted up to 75 days (WMD = 0.097/100 g).

Figure 3A shows that ADG increased (WMD = 0.048 kg/d; p < 0.001) only when FLAs were included in high-concentrate diets (>700 g/kg DM). However, the inclusion of FLAs in diets with low ( ≤ 400 g/kg DM) or moderate (401–700 g/kg DM) concentrate levels did not affect ADG (p > 0.05). Rumen pH increased (WMD = 0.336; p < 0.001) when FLAs were supplemented in diets with more than 700 g/kg DM of concentrate (Figure 3B). However, the inclusion of FLAs in diets with low (≤400 g/kg DM) or moderate (401–700 g/kg DM) concentrate levels did not affect rumen pH. Figure 3C shows that the inclusion of FLAs in diets with 401–700 g/kg DM of concentrate increased the protein content in milk (WMD = 0.089/100 g; p < 0.001). However, milk protein content was not affected when FLAs were fed in low-concentrate diets (≤400 g/kg DM).

Figure 4A shows that DMI increased (p < 0.05) when the type of FLAs used was daidzein (WMD = 0.500 kg/d), puerarin (WMD = 0.700 kg/d), and anthocyanin (WMD = 0.535 kg/d). However, DMI decreased when the type of FLAs used was naringin (WMD = −0.129 kg/d; p = 0.021) and was not affected when mixtures of FLAs were used (p > 0.05). Rumen pH increased (WMD = 0.071; p = 0.030) when FLAs mixtures were used (Figure 4B); however, it decreased when the FLAs used were daidzein (WMD = −0.350; p = 0.001) and anthocyanin (WMD = −0.100; p = 0.041). Likewise, when the type of FLAs used was naringin, the rumen pH was not affected (p > 0.05). Figure 4C shows that the serum concentration of SOD increased (p < 0.001) only when the FLAs used were daidzein (WMD = 9.373 U/mL) and puerarin (WMD = 19.733 U/mL). However, the serum SOD concentration was not affected when anthocyanin or FLA mixtures were used (p > 0.05). On the other hand, milk production increased (p < 0.001; Figure 4D) when mixtures of FLAs (WMD = 0.701 kg/d) and daidzein (WMD = 3.923 kg/d) were used; however, it decreased when the FLAs used were anthocyanins (WMD = −1.612 kg/d; p < 0.001). Figure 4E shows that milk protein content increased when mixtures of FLAs (WMD = 0.113/100 g; p < 0.001) and daidzein (WMD = 0.174/100 g; p = 0.044) were used. However, the protein content in milk was not affected when the FLAs used were anthocyanins (p > 0.05). Milk fat content increased (p < 0.01; Figure 4F) in response to supplementation with mixtures of FLAs (WMD = 0.106/100 g) and daidzein (WMD = 0.373/100 g); however, it was not affected by anthocyanin supplementation (p > 0.05).

Serum SOD concentration increased (WMD =11.016 U/mL; p < 0.001) when FLAs extracts were added to diets (Figure 5A). However, when FLAs were supplied as part of the diet ingredients, serum SOD concentration was not affected (p > 0.05). Milk production increased (WMD = 2.748 kg/d; p < 0.001) in response to supplementation with FLAs extracts (Figure 5B). However, milk production was not affected (p > 0.05) when FLAs were supplied as part of the diet ingredients.

Figure 6A shows that milk production increased when FLAs doses ≤600 mg/kg DM were used (WMD = 1.774 kg/d; p < 0.001). However, milk production decreased (WMD = −1.209 kg/d; p = 0.002) when the FLAs doses used were >600 mg/kg DM. In addition, serum SOD concentration increased regardless of the type of bovine used (p < 0.001; Figure 6B). However, the effect was greater when FLAs were offered to beef cattle (WMD = 14.712 U/mL) than dairy cows (WMD = 3.615 U/mL). Likewise, milk fat content increased (WMD = 0.217/100 g; p < 0.001) when FLAs were offered to cattle that were longer than 100 days in milk (Figure 6C). However, in cattle that were up to 100 days in milk, FLA supplementation did not affect milk fat content (p > 0.05).

It has been reported that the dietary inclusion of FLAs increases the relative abundance of ruminal bacteria involved in fiber degradation in adult sheep and cattle (20, 47). This effect could increase the rate of passage of feed particles in the rumen and result in higher DMI. In addition, in ruminants (yaks and sheep) dietary supplementation with FLAs increases the relative rumen abundance of the bacterial family Rikenelleceae (48, 49), which have a positive correlation with DMI in beef cattle (50). Therefore, similar effects of FLAs supplementation in the present meta-analysis partially explain the observed increase in DMI. On the other hand, in beef cattle, it has been reported that FLAs supplementation reduces the gene expression of bitter taste receptors (TAS2R, such as TAS2R7, TAS2R16, TAS2R38, and TAS2R39) in the rumen (18) and duodenal epithelium (51). This effect could decrease the release of anorexigenic molecules and increase DMI, since the activation of TAS2R triggers the release of anorexigenic molecules, such as cholecystokinin and peptide YY (52, 53). However, a subgroup analysis revealed that naringin supplementation decreased DMI. Naringin is part of the flavanones (a particular class of FLAs), which are abundant in citrus and impart a bitter taste (33). This effect could reduce the food's palatability and explain the lower DMI observed in response to naringin supplementation.

Previous studies (20, 47) have reported that, in adult ruminants (sheep and cattle), dietary supplementation with FLAs increases the relative abundance of ruminal bacteria of the genus Ruminococcus. Within this genus are the species Ruminococcus albus and R. flavefaciens, which play an important role in fiber degradation in the rumen (54). Kim et al. (55) observed that, under in vitro conditions, FLAs (catechins) increase the relative abundance of Fibrobacter succinogenes bacteria, which are also involved in fiber degradation in the rumen. Low doses (60 mg/kg body weight) of FLAs (mixtures of various types not reported) have been documented to increase the relative abundance of fungi in the rumen of dairy cows by up to 79% (56). According to Akin and Borneman (57), rumen fungi can completely penetrate the cell wall and produce large amounts of cellulases, hemicelluloses, and xylanases, which can increase cellulose degradation. In beef cattle, Niu et al. (58) observed that the dietary inclusion of plants with FLAs increased the relative abundance of rumen bacteria of the genus Succinivibrio, which have been positively correlated with NDFD, ADFD, and DMD in cattle (58). Furthermore, Zhao et al. (47) reported that, in growing lambs, supplementation with FLAs (anthocyanins) extracts decreases the relative abundance of ruminal microorganisms of the genus Prevotella, which have been negatively correlated with CPD in dairy cows (59). Thus, similar effects of FLAs supplementation in the present study partially explain the observed increases in CPD, NDFD, ADFD, DMD, and OMD.

In the present study, supplementation with FLAs increased DMI, CPD, NDFD, ADFD, EED, OMD, and DMD, which partially explains the higher ADG and lower FCR observed. In dairy cows, Zhan et al. (56) reported that dietary supplementation with FLAs increases the relative abundance of Tenericutes and Mollicutes rumen microorganisms, which have been positively correlated with ADG in finishing lambs (48). In growing lambs, FLAs supplementation reduces the relative ruminal abundance of the bacterial family Veillonellaceae (47), which has a negative correlation with ADG in sheep (60). Du et al. (48) reported that the dietary inclusion of plants containing FLAs increases the relative abundance of the Rikenellaceae microbial family in rumen fluid, which has a positive and negative correlation with ADG and FCR in beef cattle, respectively (50). Dorantes-Iturbide et al. (61) reported that, in finishing lambs, supplementation with low doses (1 g/kg DM) of polyherbal additives with FLAs increases up to 23% the efficiency of utilization of dietary energy for weight gain. Furthermore, supplementation with FLAs-rich plants increases muscle protein synthesis in lambs (62). Thus, similar effects of FLAs supplementation in the present study partially explain the observed increase and decrease for ADG and FCR, respectively.

In beef cattle, supplementation with FLAs (200 and 400 mg/kg DM) increases serum levels of IGF-1 (insulin-like growth factor 1) (24), which have a positive correlation (r within 0.61 and 0.67) with ADG in ruminants (63). In addition, in the present meta-analysis, higher serum concentrations of antioxidant enzymes (SOD, CAT, and GPx) and immunoglobulins (IgA, IgG, and IgM) were observed in response to FLAs supplementation. These effects could reduce oxidative stress and improve the health status of the animals, which could result in improved animal performance. On the other hand, a subgroup analysis revealed that ADG was significantly increased when FLAs were offered with high-concentrate diets (>700 g/kg DM). In beef cattle fed high-concentrate diets, FLAs supplementation increases the duodenal flux of microbial protein (64), which may increase metabolic amino acid availability and lead to higher ADG. In addition, previous studies (51, 65) have shown that the dietary inclusion of FLAs (400 mg/kg DM) improves the health of the rumen epithelium in beef cattle fed diets high in concentrate. This effect could result in increased absorption of volatile fatty acids and lead to increased ADG since the rumen epithelium contains papillae that serve as absorptive structures (66).

It has been documented that FLAs supplementation increases the number and diameter of muscle fibers in the ruminant Longissimus dorsi muscle (62, 67), which may result in increased LDMA. However, in the present study, FLAs supplementation did not affect LDMA. On the other hand, Liang et al. (25) reported that, in beef cattle, supplementation with FLAs (500 mg/kg DM) increases serum leptin concentration, which has been positively correlated with BFT in beef cattle (68). Consequently, similar effects of FLAs supplementation in the present study partially explain the higher BFT observed. In addition, FLAs promote adipogenesis in the subcutaneous adipose tissue of beef cattle through changes in the expression of several genes (delta like non-canonical notch ligand, insulin like growth factor binding protein 2, wnt family member 6, enhancer binding protein beta, DNA-binding protein inhibitor ID-3, sonic hedgehog protein, and family zinc finger 1) involved in adipogenesis differentiation of subcutaneous adipocytes (69).

According to Celi (5), the excessive accumulation of reactive oxygen species (ROS) causes oxidative stress in ruminants. Shi et al. (70) mentioned that FLAs can be used as natural antioxidants for cattle since they stimulate antioxidant enzymes and eliminate ROS. In the present study, supplementation with FLAs increased the serum levels of SOD, CAT, and GPx. These results suggest that FLAs reduce the oxidative stress caused by ROS in bovines since SOD, CAT, and GPx play an important role in converting ROS into other compounds that are less damaging to the tissues and cells of organisms (10). Furthermore, FLAs have been reported to induce activation of the transcription factor Nrf2 (71), which activates several antioxidant enzymes (10). Consequently, similar effects of FLAs consumption in the present meta-analysis partially explain the observed increases in SOD, CAT, and GPx.

In the present meta-analysis, FLAs supplementation increased TAC in beef and dairy cattle blood serum. This result suggests that the consumption of FLAs improves the total antioxidant status of bovines since TAC considers the total antioxidants present in the blood serum (5). Furthermore, Ghiselli et al. (72) mentions that serum TAC levels obtained after consuming products with antioxidants serve as indicators of the absorption and bioavailability of ingested antioxidants. Consequently, the higher TAC observed in the present study suggests that FLAs consumed by bovines may be absorbed and transferred to the bloodstream to act as blood antioxidants. Furthermore, it has been documented that TAC and ROS serum levels are negatively correlated (73). Therefore, the observed reduction of TAC in the present study suggests that FLAs supplementation decreases ROS in bovine blood serum. On the other hand, supplementation with FLAs decreased the serum concentration of MDA. This result suggests that the consumption of FLAs decreases lipid peroxidation in cattle blood because when lipid peroxidation is low, serum levels of MDA decrease (74).

According to Zhan et al. (75), immunoglobulins are a type of protein with chemical structures similar to antibodies, which participate in the regulation of immune responses. Therefore, obtaining information related to serum immunoglobulin concentrations in ruminants is important since it is an indicator of immunity against pathogenic microorganisms (76). Wolf et al. (77) mention that IgA inhibits the release of inflammatory cytokines, phagocytosis, and antibody-dependent cellular cytotoxicity. In addition, IgM and IgG act against infection since they participate in the phagocytic system and activate the complement system (24). In the present meta-analysis, FLAs supplementation increased serum IgA, IgG, and IgM concentrations, suggesting that FLAs improve immune competence in cattle. The mechanism of action of FLAs on serum immunoglobulin concentrations has not been studied in ruminants. However, FLAs have been documented to increase the expression of genes encoding IgA in mice (78). Likewise, various FLAs increase the number and activity of B1 and B2 lymphocytes (79), which secrete IgG and IgM (80, 81). Therefore, similar effects of FLAs consumption in the present meta-analysis would explain the observed increases in IgA, IgG, and IgM.

In the present meta-analysis, FLAs supplementation did not affect rumen pH. This result suggests that FLAs do not affect the stability of rumen functions in bovines since rumen pH is an important indicator of internal rumen homeostasis (36, 82). However, a subgroup analysis revealed that rumen pH increased when FLAs were offered in high-concentrate diets (>700 g/kg DM). Under in vitro conditions, FLAs decrease the concentration of lactate-producing bacteria (Streptococcus bovis) (83). In addition, in beef cattle fed high-concentrate diets, FLAs supplementation increases the abundance of lactate-consuming bacteria (Megasphera elsdenii and Selenomonas rumiantium) (64, 84). Similar effects of FLAs consumption in the present study could result in a lower rumen lactate concentration, which partially explains the increased rumen pH. On the other hand, the ruminal concentration of NH₃-N is the primary nitrogenous substrate used by rumen bacteria for microbial protein synthesis (85). Therefore, the absence of changes observed in the present study for the ruminal concentration of NH₃-N suggests that, in cattle, FLAs supplementation does not affect the synthesis of microbial protein in the rumen. Likewise, the absence of changes observed for NH₃-N suggests that FLAs supplementation does not affect the balance between rumen ammonia release and uptake.

Balcells et al. (64) mentioned that FLAs supplementation improves rumen fermentation in cattle. In the present study, supplementation with FLAs increased the rumen concentration of propionate with no effect on the concentration of acetate and butyrate. It has been reported that FLAs supplementation decreases the relative abundance of the microbial families Succiniclasticum and Christensenellaceae (47), which negatively correlates with the rumen concentration of propionate in sheep (67). In addition, under in vitro conditions, FLAs increase the relative abundance of the microbial family Succinivibrionaceae (86), which positively correlates with the rumen concentration of propionate in beef cattle (87). Therefore, similar effects of FLAs consumption in the present meta-analysis partially explain the increased ruminal propionate concentration. Furthermore, the observed increase in propionate suggests that FLAs increase energy availability for growth and production in cattle, since ruminal propionate is the main precursor of gluconeogenesis in ruminants (88).

FLAs supplementation decreased the ruminal concentration of total protozoa. This effect could improve the utilization efficiency of the protein and energy consumed by bovines since the reduction of rumen protozoa leads to less rumen protein degradation (89) and decreases enteric methane emissions (90).

The supplementation with FLAs did not affect the meat's pH or CL. These results indicate that the FLAs do not affect the quality or the water-holding capacity (WHC) of beef since the pH and CL serve as indicators to evaluate the quality (91) and WHC of the meat (92), respectively. On the other hand, lower ShF and MDA were observed in beef cattle meat in response to FLAs supplementation. These results indicate that FLAs improve beef's tenderness and oxidative stability, as ShF and MDA are indicators of meat tenderness (93) and lipid peroxidation (94), respectively. The lower ShF could be related to the reduction in IMF observed in beef from bovines supplemented with FLAs, since there is a negative correlation (r = −0.54) between ShF and IMF in beef (95). In beef cattle, low doses (400 mg/kg DM) of FLAs have been reported to decrease skeletal muscle fiber diameter (67), which is positively correlated with ShF in beef (96). The reduction observed in the present study for lipid peroxidation of meat partially explains the lower ShF, as oxidation decreases post-mortem calpain activity and myofibrillar proteolysis, leading to higher ShF (97).

The reduction observed for MDA in meat indicates that FLAs supplementation improves beef's quality and shelf life because when oxidation reactions in meat increase, the quality, and shelf-life decrease (98). Previous studies (62, 99) have reported that FLAs supplementation increases the activity of SOD, CAT, and GPx in the Longissimus dorsi muscle of small ruminants. Therefore, similar effects of FLAs consumption in the present meta-analysis partially explain the lower MDA content observed. On the other hand, it is widely documented that meat color is a crucial factor that consumers consider when choosing fresh meat (93). L^* and a^* values are related to meat brightness and metmyoglobin content, respectively (93, 100). In the present meta-analysis, supplementation with FLAs did not affect L^* and a^* in meat, indicating that FLAs do not affect metmyoglobin formation or appearance in beef. Furthermore, the lower b^* observed in response to FLAs supplementation is positive, as consumers expect to find low b^* values in fresh meat (101).

FLAs supplementation did not affect the meat's protein, moisture, and ash content; however, the IMF increased. These results indicate that FLAs do not negatively affect the nutritional value of beef since the protein and ash content of the meat are related to its nutritional value (93, 102). In contrast, the higher IMF observed could be positive since IMF correlates positively with beef's tenderness and juiciness (103). In addition, some FLAs increase the adipogenesis of bovine preadipocytes (104), which participate in the deposition of IMF (105). In pigs, FLAs supplementation increases skeletal muscle PPARγ mRNA expression levels (106), which positively correlates with IMF (107). Similar effects of FLAs consumption in the present study partially explain the higher IMF observed.

It has been mentioned that increasing the utilization efficiency of ingested feed is necessary to improve milk production in ruminants (108). In the present study, FLAs supplementation increased DMD, OMD, CPD, NDFD, ADFD, and EED. These results indicate that FLAs increase the utilization efficiency of ingested feed and partially explain the higher milk production observed in response to FLAs supplementation. In addition, the higher milk production could be related to the increased ruminal propionate concentration observed since milk production in dairy cows increases curvilinearly in response to the supply of gluconeogenic precursors (109). In lactating buffaloes, it has been reported that supplementation with FLAs-rich plants increases serum somatotropin levels by up to 50% (110), which positively correlates with milk production in dairy cows (111). It has been documented that FLAs decrease the ruminal abundance of Clostridium microorganisms (112), which negatively correlates with milk production in bovines (113). In growing sheep, FLAs supplementation increases the relative abundance of the Ruminococcaceae microbial family (47), which positively correlates with milk production in dairy cows (113). Consequently, similar effects of FLAs consumption in the present meta-analysis partially explain the higher milk production observed.

Higher protein and fat content in milk was observed in response to FLAs supplementation. Under in vitro conditions, FLAs decrease the relative abundance of Clostridium (112) and Methanobrevibacter spp. (114), which negatively correlates with the percentage of milk protein in ruminants (115, 116). In beef cattle, FLAs supplementation increases the ruminal presence of the microbial family Succinivibrionaceae (57), which positively correlates with the protein content in milk from dairy cows (116). In the present study, the FLAs decreased the ruminal concentration of total protozoa, which negatively correlates with the fat content in the milk of small ruminants (117). In dairy cows, Kong et al. (59) detected that FLAs supplementation increases the relative rumen abundance of the microbial genus Butyrivibrio, which has a positive correlation with the fat percentage in dairy cows (116). In dairy goats, FLAs supplementation increases the expression of genes involved in milk fat synthesis, such as genes related to de novo fatty acid synthesis [acetyl-CoA carboxylase α (ACACA), fatty acid synthase (FASN), and stearoyl-CoA desaturase (SCD1)] and triglyceride synthesis [diacylglycerol Oacyltransferase 1 (DGAT1), diacylglycerol O-acyltransferase 2 (DGAT2), and glycerol-3-phosphate acyltransferase 1 (GPAM)] (118). Briefly, acetyl-CoA and malonyl-CoA are condensed under FASN catalysis, two carbon atoms are added to the carboxyl of the fatty acid, the ACACA gene limits the rate of the process, and the SCD1 gene catalyzes the synthesis of monounsaturated fatty acids (118). Likewise, the GPAM gene catalyzes the acyl group transfer from acyl-CoA to generate 1-acylglycerol-3-phosphate, while the GDAT1 and DGAT2 genes catalyze the formation of triglycerides with fatty acyl-CoA (118). Therefore, similar effects of FLAs consumption in the present study partially explain the increased milk fat and protein content observed. On the other hand, FLAs supplementation did not affect the lactose content in milk. This result was not expected since the FLAs increased the rumen concentration of propionate, which is the primary short-chain fatty acid required for lactose biosynthesis (108).

Tong et al. (119) mention that SCC is a widely used indicator to assess the health of the mammary gland and the quality of milk in bovines. For example, an increase in SCC is associated with intramammary infection and negatively affects raw milk quality (120). In the present meta-analysis, lower SCC was observed in response to FLAs supplementation, indicating that FLAs improve mammary gland health and milk quality in cattle. In dairy cows, FLAs supplementation decreases the presence of Staphylococcus bacteria in milk (121), which positively correlates with SCC in ruminant milk (122). In addition, IgA has been reported to be involved in the protection of mucous membranes, IgM is the first line of defense against infections, and IgG plays an important role in the immune response against infections (75, 121). In the present meta-analysis, IgG, IgA, and IgM serum levels increased in response to FLAs supplementation. Therefore, similar effects of FLAs consumption in the present study partially explain the lower SCC observed.

The present meta-analysis was limited to research conducted only in beef cattle and dairy cows and may not apply to other ruminant species. In addition, high heterogeneity was detected in most of the response variables evaluated, which may represent a limitation in applying the global results obtained. However, this problem was diminished with the use of subgroup analysis, which allowed us to identify the specific conditions under which FLAs could be used successfully to improve different important parameters in beef cattle and dairy cows. Finally, this meta-analysis also establishes the steps for implementing future standardized experimental designs on the use of FLAs as growth promoters and natural antioxidants in beef cattle and dairy cows.