The possibility and benefits of reducing protein levels in broiler chicken diets have been discussed in other studies, although with controversial results (Attia et al., 2018; Soares et al., 2020). Reduced feed protein levels also bring a reduction in the inclusion of soybean meal, the main source of protein in field diets (Soares et al., 2020), which increases the inclusion of synthetic amino acids and corn and reduces the inclusion of soybean oil or another lipid source. However, the reduced inclusion of a lipid source may impair poultry performance mainly due to the extra-caloric effects of oils and fats.

Soares et al. (2020) found worse performance and higher caloric increment for male broilers in the growing phase with protein reduction from 22 to 16%, which can be attributed to protein content, but also to the lower percentage of soybean oil in the 16% protein diet. This suggests that in addition to nutritional levels, the concentrations of ingredients also contributed to the result, thus highlighting the extra-caloric effects of lipid sources.

Besides providing energy, the addition of lipids in diet leads to changes in gastrointestinal physiology, such as reduced peristaltic movements and improved nutrient digestibility due to increased cholecystokinin production, which stimulates gallbladder contraction and secretion of pancreatic enzymes (Ravindran et al., 2016). Thus, nutrients in the diet are under the action of enzymes and digestive juices for a longer period of time and, thus, can be better assimilated in the gastrointestinal tract. In addition, there is a reduction in caloric increase resulting from lipid metabolism (Baião and Lara, 2005). Thus, the manipulation of protein in association with lipid benefits can be explored to improve performance and carcass characteristics of broiler chickens because proteins and amino acids are closely related to muscle development (Baião and Lara, 2005; Attia et al., 2018).

In addition to diet, broiler sex is also an important factor, as higher growth rates and carcass yield are found in male broilers (Marcato et al., 2010). To meet the demand of the poultry industry for heavier carcasses, nutritional strategies and older slaughtering age have been adopted for female broilers. On the other hand, heavier carcasses are associated with increased occurrence of myopathies in chicken meat, as these changes are correlated with high carcass yield, sex, and diet, beyond chicken handling at slaughter line (Bailey et al., 2015).

The most frequent myopathies are white striping (Kuttappan et al., 2012) and wooden breast (Sihvo et al., 2014) observed in the pectoralis major muscle. Green breast or deep pectoral myopathy have also been observed in the pectoralis minor (Kijowski et al., 2009), in addition to dorsal muscle myopathy in the latissimus dorsi. These myopathies are not associated with infectious agents, yet they impair the appearance, texture, and quality of meat. Prior to the macroscopic appearance of myopathies, some histology changes, such as muscular fiber fragmentation, hyalinization, tissue necrosis, and macrophage infiltration, can be observed in tissues (Zimermann et al., 2012).

In view of the above, our objective was to evaluate the effects of two feed protein levels and types of formulation to female broilers in the growing phase on performance, metabolism, and myopathies. In addition, the effects of slaughter age on performance and myopathies were also evaluated.

The experiment was conducted in Belo Horizonte, MG, Brazil (19°55' S and 43°56' W). The current experimental protocol was approved by the Ethical Principles in Animal Experimentation Committee (case number 179/2018).

There was no interaction between diet and age for any of the performance variables studied (P>0.05). Broilers that received the 19P diet had higher FI than those that received the 21P and 21E diets (P≤0.05), while FI was intermediate for the other treatments. Regardless of diet, higher FI values were found at 44 days compared with 37 days (P≤0.05) (Table 3).

Table 3

	Performance
	Initial weight (kg)	FI (kg)	BWG (kg)	FCR	Viability (%)	PF
Diet
19P	0.937	2.930a	1.689	1.73b	100.00	499.80b
21P	0.928	2.842bc	1.705	1.66a	100.00	524.51ab
19E	0.916	2.901ab	1.720	1.68a	100.00	521.80ab
21E	0.928	2.811c	1.706	1.64a	98.71	524.45ab
19E+Aa	0.926	2.879abc	1.731	1.66a	98.71	526.14a
Age
37 days	-	2.326b	1.390b	1.67	99.48	516.84
44 days	-	3.419a	2.031a	1.68	99.48	521.84
P-value
Diet	0.5679	0.0030	0.2577	<0.0001	0.1764	0.0349
Age	-	<0.0001	<0.0001	0.3041	0.5000	0.3992
Diet × age	-	0.9898	0.7996	0.9321	-	0.9156
SEM	-	0.076	0.048	0.040	1.455	22.78

FI - feed intake; BWG - body weight gain; FCR - feed conversion ratio; PF - productive factor; SEM - standard error of the mean.

Diet did not affect broiler chicken BWG (P>0.05). Broilers had higher BWG at 44 days than at 37 days (P≤0.05). Broilers fed the 19P diet had higher FCR compared with those in the other treatments (P≤0.05), but age had no effect on FCR (P>0.05). Neither diet (P>0.05) nor age (P>0.05) had an effect on VIA. Treatment 19E+Aa showed higher PF than treatment 19P (P≤0.05), while the other treatments showed intermediate PF values. Regardless of diet, age had no effect on PF (P>0.05) (Table 3).

Values of DMDC for treatments 19P, 21P, and 19E+Aa were higher than those for treatment 21E, while treatment 19E was intermediate and similar to the others (P≤0.05). Treatment 19E+Aa had better CPDC than the other treatments (P≤0.05). Treatment 19E+Aa had lower NE than treatments 21P and 21E, while treatments 19P and 19E had intermediate values (P≤0.05). Animals that received treatments 19E, 21E, and 19E+Aa had better EEDC compared with animals receiving 19P and 21P (P≤0.05). Treatment 21P had lower AME and AMEn than the other treatments (P≤0.05) (Table 4).

Table 4

Diet	DMDC	CPDC	NE	EEDC	AME	AMEn
19P	74.58a	57.14b	99.45ab	77.25b	3,306a	3,154a
21P	73.16a	56.88b	110.78a	77.82b	3,129b	2,956b
19E	71.49ab	59.28b	95.64ab	80.80a	3,396a	3,227a
21E	69.78b	57.18b	113.55a	80.93a	3,396a	3,219a
19E+Aa	73.66a	63.13a	86.18b	81.09a	3,403a	3,221a
P-value	<0.001	0.049	0.013	0.034	<0.001	<0.001
SEM	0.3245	0.6938	2.5536	0.6208	0.1411	0.1225

DMDC - dry matter digestibility coefficient (%); CPDC - crude protein digestibility coefficient (%); NE - nitrogen excretion (g/kg); EEDC - ether extract digestibility coefficient (%); AME - apparent metabolizable energy (kcal/kg); AMEn - apparent metabolizable energy corrected for nitrogen balance (kcal/kg); SEM - standard error of the mean.

Neither diet (P>0.05) nor slaughter age (P>0.05) had an effect on the white striping score or on the appearance of wooden breast in female broilers (Table 5). For the pectoralis major muscle (Figures 1A and 1B), connective tissue intensity was lower for the 19P treatment compared with the other treatments (P≤0.05). The other histological variables were not affected by diet (P>0.05). Age had no effect on any of the evaluated variables in this muscle (P>0.05) (Table 6).

Table 5

	White striping1	Wooden breast2
	Score	Absent (%)	Present (%)
Diet
19P	0.91	62.50	37.50
21P	0.81	54.55	45.45
19E	0.95	58.33	41.67
21E	1.33	37.50	62.50
19E+Aa	0.72	63.64	36.36
Age
38 days	0.83	53.33	46.64
46 days	1.03	57.14	42.46
P-value
Diet	0.4383	0.368
Age	0.4565	0.680

¹ Means compared by the Kruskal Wallis test (P<0.05) for the independent variable and the Wilcoxon test (P<0.05) for the dependent variable.

Table 6

	Variable
	MFN	HD	FIB	ADP	LHI	LS
Diet	Pectoralis major
19P	5.10	15.15	0.00	0.40	0.70	1.50
21P	6.17	15.69	0.25	0.62	0.87	1.75
19E	5.58	17.14	0.11	0.44	1.11	1.67
21E	5.95	12.69	0.25	0.50	0.62	1.62
19E+Aa	5.64	25.09	0.45	0.54	0.82	1.64
Age
38 days	6.58	16.68	0.12	0.37	0.75	1.79
46 days	4.66	18.56	0.32	0.64	0.91	1.45
P-value
Diet	0.879	0.507	0.138	0.897	0.183	0.922
Age	0.057	0.349	0.116	0.079	0.187	0.095
	Pectoralis minor
Diet
19P	2.52	13.00	0.25	0.37	0.75	1.00
21P	4.97	21.55	0.33	0.00	1.00	1.55
19E	3.66	15.60	0.20	0.10	0.70	1.22
21E	3.17	16.94	0.12	0.12	0.75	1.12
19E+Aa	4.96	22.44	0.62	0.50	1.12	1.75
Age
38 days	2.27b	12.28b	0.13	0.13	0.65b	0.87b
46 days	5.71a	24.32a	0.50	0.30	1.10a	1.89a
P-value
Diet	0.301	0.563	0.607	0.164	0.392	0.423
Age	0.001	0.002	0.076	0.296	0.004	>0.001
	Latissimus dorsi
Diet
19P	2.78	14.58	0.45	1.00	0.54	1.00
21P	2.52	15.83	0.40	0.90	0.60	0.90
19E	2.46	20.64	0.33	0.89	0.44	0.89
21E	3.08	16.85	0.20	1.20	0.70	1.10
19E+Aa	2.36	17.69	0.20	0.90	0.40	0.60
Age
38 days	2.74	15.63	0.13b	0.56b	0.22b	0.83
46 days	2.57	18.16	0.48a	1.33a	0.81a	0.96
P-value
Diet	0.650	0.494	0.817	0.947	0.972	0.488
Age	0.868	0.262	0.024	0.003	>0.001	0.425

MFN - muscle fiber necrosis; HD - hyaline degeneration; FIB - fibrosis; ADP - adipocytes; LHI - lymphohistiocytic infiltration; LS - lesion score.

For the pectoralis minor muscle (Figure 1C), diet had no effect on any of the evaluated parameters (P>0.05). However, there was an increase in the intensity of necrosis, hyaline degeneration, lymphohistiocytic infiltration, and lesion score at 46 days compared with 38 days (P≤0.05). Age had no effect on fibrosis and adipocyte intensity (P>0.05) (Table 6).

Figure 1 A - Normal pectoralis major muscle; B - Pectoralis major muscle with fiber degeneration (→), fiber necrosis (▲), and inflamatory inflitrate (★); C - Pectoralis minor muscle with fiber degeneration (→), fiber necrosis (▲), and inflamatory inflitrate (★); D - Latissimus dorsi muscle with fiber degeneration (→) and fiber necrosis (▲). Hematoxylin and eosin.

For the latissimus dorsi muscle (Figure 1D), diet had no effect on any of the variables studied (P>0.05). However, higher fibrosis, adipocytes and lymphohistiocytic infiltration were more frequent at 46 days than at 38 days (P≤0.05). Age had no effect on necrosis, hyaline degeneration, and lesion scores in this muscle (P>0.05) (Table 6).

Despite numerous studies aiming to reduce dietary protein in broiler feed (Si et al., 2004; Leeson and Summers, 2005; Attia et al., 2018; Soares et al., 2020), studies considering protein reduction without altering the concentration of lipids, fiber, and starch in the diet cannot be found. These factors may influence nutrient utilization and, consequently, animal performance.

The higher FCR observed for the 19P treatment may be associated with the lower inclusion of oil in the diet since the 19E treatment, which had the lowest FCR, had the same CP content but with higher oil inclusion. Inclusion of lipids in poultry diets improves FCR because lipids reduce gastric movement, slowing the passage of diet in the digestive tract (Baião and Lara, 2005). This, in turn, improves nutrient utilization and increases the feeling of satiety, and thus lipids are known as gastric emptying inhibitors (Khoddami et al., 2018).

In the present study, the experimental feed represented by treatments 19E, 21E, and 19E+Aa, designed to isolate the lipid, fiber, and starch factors, which can affect nutrient utilization, did not affect BWG and FCR. When these factors are isolated, it is possible to perceive the beneficial effects of lipids on broiler performance because for treatments 19P and 21P, in which these factors were not isolated, the former, with lower lipid content, had worse FCR.

In a study using different CP levels (16, 18, 20, or 22%) for male broilers, Soares et al. (2020) observed higher BWG and lowest FCR for the two highest CP levels, which was explained by CP level and inclusion of oil in diets. Similarly, Attia et al. (2018) evaluated a 2% increase in CP contents, in addition to energy increase along with the increase of CP in the initial diets, and growth of broilers under heat stress. The authors found, along with the energy increase, higher BWG and lower FCR for the treatment with higher CP content (22%), whose formulation had about 2% greater inclusion of oil.

According to Cobb (2018), the BWG of female broilers from 21 to 37 days is 1.332 kg and from 21 to 44 days, it is 1.932 kg, while FCR is 1.79 from 21 to 37 days and 1.92 from 21 to 44 days. However, the present study found higher BWG (1.390 kg from 21 to 37 days and 2.031 kg from 21 to 44 days) and lower FCR (1.67 from 21 to 37 days and 1.68 from 21 to 44 days). Thus, FCR did not worsen with increasing age as suggested by Cobb (2018), as the increase in FI was proportional to the increase in BWG observed at 37 and 44 days. This proportional increase may have been achieved by adequate conditions in the experimental environment, which was also reflected in the low mortality rate, leading to similar viability results at 37 and 44 days. The higher PF value of treatment 19E+Aa compared with treatment 19P occurred in response to the BWG and dietary FCR found for these treatments.

The results found for performance in the present study are in agreement with nutrient digestibility and diet energy values. Reduced nutritional levels in a diet improves nutrient digestibility for broiler chickens (Viola et al., 2008); however, the lack of standardization of the nutritional matrix of corn gluten meal (Santos, 2004) and higher inclusion of soybean meal may have worsened DMDC for the 21E treatment, which was similar to that of 19E and lower than the others. The amount and quality of carbohydrates, proteins, and fats contained in different diets can alter the retention time of digestion and thus influence the efficiency of digestion and absorption of nutrients (Soares et al., 2020).

The greater inclusion of synthetic amino acids in the 19E+Aa treatment improved CPDC because these amino acids are absorbed faster and more efficiently than those contained in the intact dietary protein. In addition, the NE values, and also the lowest AME and AMEn values, observed in the 21P treatment confirm that the CP supply may be above the requirements. Excess CP is catabolized and excreted as uric acid, which has a high energy cost to the animal and may be associated with the loss of about 0.7 kcal of energy in the urine for each gram of dietary CP. Thus, the energy of productive processes is converted to nitrogen excretion (Noblet and Perez, 1993).

A reduction in AME and AMEn values was not observed for the 21E treatment, as was observed for 21P. This is due to the higher inclusion of oil in the diet of the 21E treatment compared with the 21P treatment, evidencing the extra-caloric effect of oil in improving energy utilization by broilers. Similarly, higher oil inclusion improved EEDC in treatments 19E, 21E, and 19E+Aa compared with the others. Increased yield of carcass and parts is one of the main objectives of the poultry industry because this factor is responsible for higher profitability of the activity (Nortey et al., 2019). The difference in protein deposition at different ages of broilers is mainly due to changes in metabolism, because the difference between anabolic and catabolic processes decreases with advancing age (Radaelli et al., 2017). When the animal reaches its peak growth, the tendency is for fat deposition to increase and protein deposition to decrease.

Linked to older slaughter age and higher carcass yields, the pursuit of rapid growth and yield of poultry has implications for animal physiology and, consequently, changes in meat quality (Bailey et al., 2015; Petracci et al., 2019). Since pectoral muscle growth is achieved by muscle fiber hypertrophy, which results in increased fiber diameter, there is also reduced space for connective tissue and, as a consequence, increased muscle degeneration. These morphological changes limit blood supply, compromising nutrient supply and the removal of metabolites produced by muscle fibers, thereby altering meat quality (Sandercock et al., 2009) and causing the onset of myopathies.

Mudalal et al. (2015) found a strong association between the occurrence of wooden breast and higher weight and white striping of the pectoralis major muscle; however, the myopathies studied here (white striping and wooden breast) were not associated with the diets provided to broilers nor with slaughter age. Furthermore, it was not possible to notice any difference in the intensity of myopathic lesions due to slaughter age.

Although the white striping myopathy has no defined etiology, it seems to be related to high growth rates, and so males have a higher incidence, which can be explained by the greater body weight of these animals (Kuttappan et al., 2012). Wooden breast in the pectoralis major of broiler chickens may be associated with white striping myopathy in the most severe cases (Sihvo et al., 2014), but its etiology and the factors that cause its appearance have not been identified, being predominantly described, as well as white striping in fast growing strains.

Histological changes in muscle tissue can be seen in broilers affected by myopathies before macroscopic perception. Changes in muscle tissues can present pathological aspects and usually affect the pectoralis major, pectoralis minor, and latissimus dorsi (Ecco and Braga, 2015). The pectoral muscles act synergistically for wing movement, with the pectoralis major as the main thruster muscle, being composed of fast-twitch myofibers with glycolytic activity, the fibers of which respond rapidly with increased white muscle mass when given high nutritional diets (Roy et al., 2006). However, although tissue lesions were observed, no association was observed between the experimental diets of the present study and histological characterization.

The connective tissue of the pectoral muscle has thousands of cells with tissue connection, support, and filling functions. Despite its regenerative capacity, when damage to skeletal muscle leads to cell destruction, connective tissue proliferation occurs, and when the muscle area is replaced by fibrous tissue, complete regeneration is inhibited. After muscle fiber degeneration, revascularization of the injured area occurs and inflammatory cells are activated to remove necrotic tissue, which requires the formation of healing connective tissue to keep the ends of the myofibrils connected (Cruz et al., 2017; Petracci et al., 2019).

These principles may underlie the lower connective tissue intensity observed for the 19P treatment, which contained lower CP content and oil inclusion, and possibly led to slower muscle development, compared with the other treatments. The changes observed in the pectoralis minor and latissimus dorsi due to slaughter age were not associated with infectious conditions. However, histological changes were possibly a consequence of ischemia and the rapid growth of these muscles (Velleman et al., 2014).

Whereas white striping and wooden breast myopathies may exhibit similar histological lesions, including myodegeneration and regeneration, fibrosis, and lipidosis (Sihvo et al., 2017; Soglia et al., 2016; Radaelli et al., 2017; Petracci et al., 2019), there is probably some common mechanism that allows both myopathies to involve the same muscle, which can also be observed in the present study. Moreover, in this study, the histological features found were consistent with macroscopic evaluation (Baldi et al., 2018).

Diet with 19% crude protein, using only commercially available ingredients, was not adequate because it worsens feed conversion. The diet with 21% crude protein, using only commercially available ingredients, reduces the availability of apparent metabolizable energy and apparent metabolizable energy corrected for nitrogen balance due to high nitrogen excretion, so it should be avoided. The experimental diet 19% crude protein plus synthetic amino acids (19E+Aa) improves the crude protein digestibility coefficient, exhibits low nitrogen excretion, and has feed conversion similar to diets with 21% crude protein (using only commercially available ingredients), 19% crude protein (using experimental ingredients), and 21% crude protein (using experimental ingredients), all with higher oil inclusion. However, the choice of the most appropriate diet should be based on the cost and availability of ingredients. In addition, slaughter of 46-day-old instead of 38-day-old female broiler chickens is recommended. Broilers slaughtered at 46 days have higher weight gain without worsening feed conversion and the intensity of myopathies.