Association of glucose metabolism and insulin resistance with feed efficiency and production traits of finishing beef steers

Abstract

Increasing nutrient utilization efficiency is an important component of enhancing the sustainability of beef cattle production. The objective of this experiment was to determine the association of glucose metabolism and insulin resistance with dry matter intake (DMI), average daily gain (ADG), gain:feed ratio (G:F), and residual feed intake (RFI). Steers (n = 54; initial body weight = 518 ± 27.0 kg) were subjected to an intravenous glucose tolerance test (IVGTT) where glucose was dosed through a jugular catheter and serial blood samples were collected. Three days after the last group’s IVGTT, steers began a 63-d DMI and ADG test. Body weight was measured on days 0, 1, 21, 42, 62, and 63, and DMI was measured using an Insentec Roughage Intake Control system (Hokofarm Group, Emmeloord, the Netherlands). To examine relationships between DMI, ADG, G:F, and RFI with IVGTT measurements, Pearson correlations were calculated using Proc Corr of SAS 9.4 (SAS Inst. Inc., Cary, NC). Additionally, cattle were classified based on DMI, ADG, RFI, and G:F, where the medium classification was set as mean ± 0.5 SD, the low classification was < 0.5 SD from the mean, and the high classification was > 0.5 SD from the mean. No associations between DMI and IVGTT parameters were observed, and no differences were detected when classifying cattle as having low, medium, or high DMI. Peak insulin concentration in response to the IVGTT tended to be correlated with ADG (r = 0.28; P = 0.07), indicating cattle with greater ADG tend to have a greater insulin release in response to glucose. Glucose nadir concentrations tended to be positively correlated with ADG (r = 0.26; P = 0.10). Additionally, the glucose nadir was greater in high-ADG steers (P = 0.003). The association of greater glucose nadir with high-ADG could indicate that high-ADG steers do not clear glucose as efficiently as low-ADG steers, potentially indicating increased insulin resistance. Further, RFI was not correlated with IVGTT measurements, but low RFI steers had a greater peak glucose concentration (P = 0.040) and tended to have a greater glucose area under the curve (P = 0.09). G:F was correlated with glucose area under the curve (r = 0.33; P = 0.050), glucose nadir (r = 0.35; P = 0.011), and insulin time to peak (r = 0.39; P = 0.010). These results indicate that glucose metabolism and insulin signaling are associated with growth and efficiency, but the molecular mechanisms that drive these effects need to be elucidated.

Insulin resistance is associated with improved feed efficiency of finishing beef cattle.

Introduction

Ensuring the sustainability of the beef production system in the United States is a growing concern for scientists and producers (Harmon, 2020), which requires a reduced environmental impact while meeting the growing global demand for protein (Hume et al., 2011). The efficient conversion of feed nutrients to human edible products is an important component of the sustainability of the U.S. beef production system. As the conversion of nutrients from feed into edible carcass components increases, less carbon and nitrogen will be excreted into the environment, thereby reducing environmental impacts. Further research is required to understand animal nutrient requirements and the potential of improving genetic selection and diet formulation to improve feed efficiency. Importantly, there is a gap in the knowledge of the underlying metabolic processes that regulate animal growth performance and efficiency, which is critical for improving the sustainability of beef production.

Glucose is a vital energy substrate for muscle and adipose tissues (Rhoades et al., 2007), with insulin being an important regulator of glucose production and utilization. Insulin influences overall nutrient partitioning as well as protein and adipose accrual by activating pathways to facilitate glucose uptake in muscle and adipose tissue, increasing energy storage (Terlouw et al., 2021). Insulin resistance is a phenomenon that occurs in many species and is defined as a state when normal concentrations of insulin produce a less-than-normal biological response (Kahn, 1978). Insulin resistance results from changes in insulin sensitivity (concentration of insulin required to produce the half-maximal response) and insulin responsiveness (maximum response to insulin). In animals, it is difficult to distinguish between alterations in insulin sensitivity and responsiveness (Petersen and Shulman, 2018); therefore, it is becoming more common for scientists to simply focus on insulin resistance. While the complex hyperinsulinemic-euglycemic clamp model is the standard for evaluating insulin resistance, it is difficult to perform on large numbers of animals. Surrogate indices such as the homeostasis model of insulin resistance (HOMA-IR; Matthews et al. 1985) and the quantitative insulin sensitivity check index (QUICKI; Katz et al. 2000) have been developed for use in humans and rely on fasting insulin and glucose concentrations. These indices have been used in ruminants, yet their utility has been questioned due to the differences in insulin actions between humans and ruminants as well as the time required to reach a fasting state in cattle (De Koster et al., 2016). The intravenous glucose tolerance test (IVGTT) is more straightforward than the hyperinsulinemic-euglycemic clamp and provides more useful data than the HOMA-IR and QUICKI values. Data from an IVGTT can show alterations in glucose clearance and alterations in the insulin response to glucose, as well as provide fasting values for insulin and glucose. Taken together, data from IVGTT can provide valuable data on insulin resistance and glucose metabolism more comprehensively than single blood samples.

Previous research demonstrates that as cattle are on a high-concentrate ratio for longer feeding periods, they develop insulin resistance (Joy et al., 2017; de Sousa et al., 2022). This is also associated with decreased growth rate and efficiency because of the confounding nature of growth curves, time on feed, and the development of insulin resistance. It is unclear if there is a causative relationship between insulin resistance and changes in growth and efficiency. Fitzsimons et al. (2014) studies show no relationship between residual feed intake (RFI) and the glucose and insulin response to an IVGTT. However, the association of glucose and insulin responses with dry matter intake and average daily gain (ADG) that contribute to RFI were not reported. Furthermore, this previous experiment utilized cattle fed primarily grass silage diets (1.74 to 1.80 Mcal of net energy for maintenance [Nem] per kg of dry matter). It is possible that RFI could be associated with cattle’s response to an IVGTT in a high-concentrate finishing ration. Despite the IVGTT response not being associated with RFI, there were positive associations of insulin receptor and protein kinase C-alpha gene expression in muscle with ADG (Fitzsimons et al., 2014). This finding supports the idea that variation in components of insulin signaling in muscle is an important contributor to observed variation in growth.

To the best of our knowledge, there has been no investigation into how the response of finishing beef cattle to IVGTT relates to DMI, ADG, feed efficiency, and RFI. Accordingly, the objective of this experiment was to determine the association of glucose metabolism and insulin resistance measured using an IVGTT early in the finishing phase with production and efficiency traits in finishing the beef cattle. We hypothesized that the response of the steers to the IVGTT, which is representative of their metabolic status entering the finishing phase, would be related to their production and efficiency traits during the finishing period.

Materials and Methods

All animal procedures were approved by the Oklahoma State University (OSU) Institutional Animal Care and Use Committee (Protocol #19-77).

Animals

Angus calves from an OSU herd were weaned at approximately 6 mo of age, transported to the Willard Sparks Beef Research Center, and placed on a receiving ration. While in the feedlot, cattle were acclimated to close human contact and trained to halters. Steers (n = 54; initial body weight [BW] = 518 ± 27 kg) were housed in two pens containing an Insentec Roughage Intake Control System (Hokofarm Group, Emmeloord, the Netherlands). Each pen contained six feed bunks designed to measure individual feed intake, and steers had access to all bunks. Bunks were managed to have approximately 1 kg of orts per day to minimize sorting. After 1 mo of the receiving ration and training to the Insentec system, cattle were transitioned to a finishing diet (DM basis: 62% dry-rolled corn, 20% wet corn gluten feed, 8% prairie hay, 5% vitamin/mineral mix, and 5% liquid supplement). The diet was formulated to meet or exceed nutrient requirements (National Academies of Sciences, 2016). Transition to the finishing diet was done over 3 wk using a two-ration blend system. Briefly, cattle were fed 75:25 starter-to-finisher ratio for the first week of transition, 50:50 for week two, 25:75 for week three, and were then fed 100% finisher for the remainder of the experiment. After 14 d of consuming the finishing ration, 5 blocks of 9 or 10 steers (blocked by BW) were moved to the OSU Animal Nutrition and Physiology Research Center and placed in individual pens for an IVGTT.

Glucose tolerance test

An IVGTT was conducted on each of the five blocks following a 12-h fast (Rathert-Williams et al., 2021, 2023) on either days −7, −6, −5, −4, and −3 of the experimental period. Conducting the IVGTT before the feeding period was chosen to determine the association with glucose metabolism and insulin resistance entering the finishing phase with feed efficiency traits. Briefly, a temporary indwelling jugular catheter (16-gauge × 13 cm; Jorgensen Labs, Loveland, CO) was placed in each steer about 1 h prior to sampling with a 76.2 to 86.4 cm catheter extension set (Oasis, Mettawa, IL). A 2.78 M glucose solution was infused at a continuous rate via a jugular catheter to provide 7.57 mmol/kg BW^0.75 over a 2-min period. Blood samples were collected at −10, 5, 10, 15, 20, 25, 30, 45, 60, 90, and 120 min relative to the glucose infusion (9 mL neutral Sarstedt Monovette, Sarstedt AG & Co. KG, Nümbrecht, Germany) with K₂EDTA added at 1.5 mg/mL and immediately placed on ice. Catheters were flushed with 10 mL of heparinized physiological saline (10 IU/mL; Thermo Fisher Scientific Chemicals, Inc., Ward Hill, MA) immediately after each blood collection. Blood samples were immediately placed in ice and centrifuged for 20 min at 3,000 × g and 4 °C. Plasma was collected and stored at −20 °C in 2 mL aliquots until further analysis. Plasma glucose was analyzed using the YSI Biochemistry Analyzer 2,900 (YSI Inc., Yellow Springs, OH) using glucose oxidase in an immobilized membrane. Plasma insulin was analyzed using a commercially available porcine insulin radioimmunoassay kit (Millipore Corporation, Billerica, MA) with insulin from bovine pancreas (Sigma-Aldrich Inc., St. Louis, MO) used to construct a standard curve (Rathert-Williams et al., 2023). The intra- and inter-assay CV were 2.55% and 7.73%, respectively.

Feed intake and growth trial

Following the IVGTT, steers were transported back to the Willard Sparks Beef Research Center, placed in the pen they were removed from previously, and DMI was measured for 63 d. Using the data from the Insentec system, a minimum meal size was set at 0.2 kg, and all feeding events within a 10-min period were considered a single meal. BW was measured on days 0, 1, 21, 42, 62, and 63. A quadratic polynomial equation was used to regress BW on days of study for each steer (Foote et al., 2015) with the slope as ADG. The ratio of ADG:DMI (i.e., gain:feed; G:F) was calculated. RFI was calculated by regressing observed DMI against ADG and metabolic body weight at the midpoint of the experiment (Foote et al., 2015). Following completion of the DMI and growth trial, cattle were shipped to a commercial slaughter facility and carcass data were collected by plant personnel.

Calculations

Data from the IVGTT were analyzed using GraphPad Prism 8.4.3 (GraphPad Software, San Diego, CA) to determine the area under the curve (AUC) for glucose and insulin. This data was also modeled as an exponential one-phase decay to calculate glucose and insulin peak, glucose clearance rate, glucose half-life, and glucose nadir in GraphPad Prism (GraphPad Software).

Variables from the IVGTT were analyzed for Pearson correlation with DMI, ADG, RFI, and G:F using the CORR procedure of SAS 9.4 (SAS Inst. Inc.). Additionally, individual animals were classified as low, medium, or high for DMI, ADG, RFI, and G:F. The Medium classification of each trait was set as the mean ± 0.5 SD; the low classification was < 0.5 SD from the mean; and the high classification was > 0.5 SD from the mean. The effect of classification on the IVGTT parameters was analyzed in a mixed model analysis using the MIXED procedure of SAS 9.4 (SAS Inst. Inc.) with the fixed effect of classification. Means were separated using the LS-Means statement with a Tukey-Kramer post hoc adjustment. Means presented are least-square means ± SEM. Significance was deemed when P ≤ 0.05 and considered a tendency when 0.05 < P ≤ 0.10. P-values are reported with two decimal places unless P < 0.05 and they are reported to three decimal places.

Results

The descriptive statistics (mean and SD) for the measured variables are shown in Table 1. ADG was positively correlated glucose nadir (r = 0.31; P = 0.025; Table 2; P-values reported in Supplementary Table S1), glucose half-life (r = −0.28; P = 0.047), time to insulin peak (r = 0.37; P = 0.007), day 21 plasma glucose concentration (r = 0.35; P = 0.020), and day 63 plasma insulin (r = 0.50; P < 0.001). Additionally, ADG tended to be positively correlated with peak insulin concentration (r = 0.26; P = 0.07) during the IVGTT. RFI was negatively correlated with glucose AUC (r = −0.29; P = 0.040) and peak glucose concentration (r = −0.31; P = 0.024) and tended to be negatively correlated with day 0 blood glucose concentration (r = −0.27; P = 0.08). G:F was positively correlated with glucose AUC (r = 0.34; P = 0.015), peak glucose concentration (r = 0.29; P = 0.036), glucose nadir (r = 0.34; P = 0.016), and time to peak insulin (r = 0.37; P = 0.007) during the IVGTT. Additionally, G:F was positively correlated to day 0 (r = 0.38; P = 0.017), 21 (r = 0.40; P = 0.011), and 63 (r = 0.34; P = 0.031) plasma glucose concentrations, and day 63 insulin concentrations (r = 0.42; P = 0.003). G:F also tended to be positively correlated with day 42 glucose concentrations (r = 0.27; P = 0.10).

Table 1.

Descriptive statistics of production traits and intravenous glucose tolerance test response of finishing beef steers¹

Item	Mean	SD
Initial body weight, kg	518	27.0
Average daily gain, kg/d	1.51	0.253
Dry matter intake, kg/d	12.2	1.24
Residual feed intake, kg/d	−0.01	0.978
G:F, kg/kg2	0.12	0.018
USDA yield grade3	3.39	0.524
Marbling score4	594	101.4
12th rib fat thickness, cm	1.76	0.377
Ribeye area, cm2	35.7	2.59
Baseline glucose, mg/dL	106	19.1
Glucose AUC2, mg/dL × min	21,852	3,218.7
Peak glucose, mg/dL	295	36.1
Glucose clearance rate, min−1	0.032	0.0337
Glucose nadir, mg/dL	109.1	29.44
Glucose half-life, min	27.0	8.86
Baseline insulin, ng/mL	0.698	0.3167
Insulin AUC2, ng/mL × min	186.1	64.22
Peak insulin, ng/mL	2.77	0.911
Insulin time to peak, min	12.8	3.78
QUICKI2	0.558	0.0710
HOMA-IR2	73.9	34.87

¹ n = 54.

²G:F, gain to feed ratio; AUC, area under the curve; QUICKI, quantitative insulin sensitivity check index; HOMA-IR, homeostasis model of insulin resistance.

³Calculated according to USDA (1997).

⁴Small⁰⁰ = 400; Modest⁰⁰ = 500; Moderate⁰⁰ = 600.

Table 2.

Pearson correlation coefficients for production traits with parameters from an intravenous glucose tolerance test performed prior to a 63-feed intake and growth measurement period of finishing beef steers¹

Item	ADG2, kg/d	DMI2, kg/d	RFI2, kg/d	G:F2, kg/kg	HCW2, kg	Fat2, cm	REA2, cm2	YG2	MS2
Baseline glucose, mg/dL	0.20	−0.03	−0.17	0.25*	0.06	−0.10	0.17	−0.16	−0.06
Glucose AUC2, mg/dL × min	0.20	−0.16	−0.29**	0.34**	0.09	−0.14	0.15	−0.16	−0.32**
Peak Glucose, mg/dL	0.12	−0.21	−0.31**	0.29**	0.06	−0.16	0.11	−0.15	−0.28*
Glucose clearance rate, min−1	0.16	0.14	0.13	0.07	−0.04	0.37**	−0.26*	0.41**	−0.27*
Glucose nadir, mg/dL	0.31**	0.05	−0.07	0.34**	0.10	−0.07	0.10	−0.08	−0.39**
Glucose half-life, min	−0.28**	−0.12	−0.04	−0.25*	0.05	−0.19	−0.01	−0.11	0.32**
Baseline insulin, ng/mL	0.16	0.15	−0.01	0.12	0.34**	−0.10	0.31**	−0.15	0.06
Insulin AUC2, ng/mL × min	0.19	0.14	−0.05	0.13	0.26*	0.01	0.34**	−0.15	0.07
Insulin peak, ng/mL	0.26*	0.19	−0.02	0.17	0.26*	−0.05	0.25*	−0.10	0.12
Insulin time to peak, min	0.37**	0.06	−0.15	0.37**	0.16	−0.05	0.07	−0.03	0.01
QUICKI2	−0.22	−0.10	0.08	−0.20	−0.27*	0.11	−0.26*	0.15	−0.03
HOMA-IR2	0.17	0.12	−0.04	0.13	0.31**	−0.13	0.39**	−0.24*	0.07
Day 0 glucose, mg/dL	0.22	−0.15	−0.27*	0.38**	0.03	−0.23	0.26*	−0.31**	−0.25*
Day 21 glucose, mg/dL	0.35**	0.06	−0.08	0.40**	0.14	−0.28*	0.26*	−0.30**	−0.33**
Day 42 glucose, mg/dL	0.17	−0.08	−0.20	0.27*	0.07	−0.20	0.32**	−0.31	−0.09
Day 63 glucose, mg/dL	0.25*	−0.03	−0.15	0.34**	−0.28*	−0.23	0.26*	−0.28*	−0.34**
Day 0 insulin, ng/mL	0.03	0.19	0.23	−0.09	0.18	0.02	0.36**	−0.16	0.08
Day 21 insulin, ng/mL	0.05	0.25	0.20	−0.13	0.34**	−0.15	0.47**	−0.30**	−0.21
Day 42 insulin, ng/mL	0.21	0.10	−0.03	0.16	0.20	−0.29**	0.34**	−0.37**	−0.16
Day 63 insulin, ng/mL	0.50**	0.14	−0.03	0.42**	0.50**	−0.35**	0.37**	−0.33**	−0.09

¹ n = 54.

²ADG, average daily gain; DMI, dry matter intake; G:F, gain to feed ratio; RFI, residual feed intake; HCW, hot carcass weight; Fat, 12th rib fat thickness; REA, ribeye area, YG, USDA yield grade; MS, marbling score; AUC, area under the curve; QUICKI, quantitative insulin sensitivity check index; HOMA-IR, homeostasis model of insulin resistance.

^*0.05 ≤ P ≤ 0.10.

^** P < 0.05.

Carcass characteristics displayed a number of correlations with IVGTT variables and glucose and insulin concentrations during the experiment. Glucose clearance rate was correlated with 12th rib fat thickness (r = 0.37; P = 0.011) and USDA YG (r = 0.41; P = 0.004) and tended to be correlated to REA (r = −0.26; P = 0.08) and marbling score (r = −0.27; P = 0.06). Hot carcass weight (HCW) was positively correlated with baseline insulin concentrations (r = 0.34; P = 0.019), HOMA-IR (r = 0.31; P = 0.034), day 21 insulin (r = 0.34; P = 0.018), and day 63 insulin (r = 0.50; P < 0.001). Ribeye area was positively correlated to baseline insulin (r = 0.31; P = 0.031), insulin AUC (r = 0.34; P = 0.018), HOMA-IR (r = 0.39; P = 0.006), day 42 glucose (r = 0.32; P = 0.033), and insulin concentrations from all collection days (r > 0.34; P < 0.018).

Classification of the steers based on DMI resulted from differences in initial BW (Supplementary Table S2; P = 0.010), ADG (P = 0.002), DMI (P < 0.001), and RFI (P < 0.001), but not G:F (P = 0.25). There was no (P ≥ 0.12; Table 3) effect of DMI classification on glucose or insulin responses to the IVGTT. The DMI classification influenced HCW (P = 0.001), so that the high DMI group had 8% greater HCW than the low DMI (P = 0.010) and the medium DMI group was intermediate and did not differ from the other two groups (P ≥ 0.20).

Table 3.

Effect of classification based on DMI and ADG on glucose and insulin response to an intravenous glucose tolerance test performed before the feed intake and growth measurement period and carcass characteristics

	Classification1
Item	Low	Medium	High	SEM2	P-value
DMI classification
n	18	17	19	—	—
Baseline glucose, mg/dL	111	105	105	5.1	0.59
Glucose AUC3, mg/dL × min	22,608	21,763	21,251	807.9	0.46
Peak glucose, mg/dL	304	297	285	9.0	0.28
Glucose clearance rate, min−1	0.027	0.026	0.029	0.0019	0.50
Glucose nadir, mg/dL	107.7	107.6	108.6	7.56	0.99
Glucose half-life, min	27.6	29.0	26.0	2.15	0.58
Baseline insulin, ng/mL	0.67	0.66	0.77	0.075	0.50
Insulin AUC, ng/mL × min	173	180	204	16.0	0.33
Insulin peak, ng/mL	2.49	2.74	3.04	0.225	0.20
Insulin time to peak, min	13.8	11.3	13.3	0.92	0.12
QUICKI3	0.55	0.58	0.55	0.018	0.46
HOMA-IR3	75.7	65.4	81.1	9.03	0.44
Hot carcass weight, kg	403b	417ab	436a	6.0	0.001
12th rib fat thickness, cm	1.76	1.84	1.71	0.109	0.65
Ribeye area, cm2	90.2	90.7	89.9	1.81	0.95
USDA yield grade4	3.30	3.46	3.49	0.148	0.60
Marbling score5	601	615	581	28.5	0.67
ADG classification
n	18	17	19	—	—
Baseline glucose, mg/dL	103	106	111	4.7	0.40
Glucose AUC3, mg/dL × min	21,000	21,486	23,101	796.0	0.14
Peak glucose, mg/dL	289	287	309	8.9	0.16
Glucose clearance rate, min−1	0.026	0.026	0.030	0.0018	0.20
Glucose nadir, mg/dL	99.3b	99.5b	124.7a	6.63	0.010
Glucose half-life, min	28.69	28.40	25.36	2.066	0.44
Baseline insulin, ng/mL	0.62	0.74	0.75	0.080	0.41
Insulin AUC3, ng/mL × min	171	197	191	16.1	0.46
Insulin peak, ng/mL	2.57	2.82	2.92	0.229	0.52
Insulin time to peak, min	12.22ab	11.56b	14.71a	0.899	0.036
QUICKI3	0.58	0.55	0.54	0.018	0.22
HOMA-IR3	64.4	80.4	79.6	8.68	0.32
Hot carcass weight, kg	402b	423a	434a	6.1	0.001
12th rib fat thickness, cm	1.71	1.97	1.66	0.103	0.07
Ribeye area, cm2	89.5	90.7	90.5	1.80	0.86
USDA yield grade4	3.26	3.63	3.40	0.144	0.18
Marbling score5	602	595	595	28.8	0.97

¹Steers ± 0.5 SD of the mean for either DMI or ADG were classified as medium. Those with DMI or ADG > 0.5 SD from the mean were high and those < 0.5 SD from the mean were low.

²Standard error of the mean. The most conservative (i.e., largest) estimate of SEM is reported for each item.

³AUC, area under the curve; QUICKI, quantitative insulin sensitivity check index; HOMA-IR, homeostasis model of insulin resistance.

⁴Calculated according to USDA (1997).

⁵Small⁰⁰ = 400; Modest⁰⁰ = 500; Moderate⁰⁰ = 600.

^abcSuperscript letters mean differ (P ≤ 0.05) for main effect of classification.

DMI, dry matter intake; ADG, average daily gain.

The initial BW of steers classified based on ADG did not differ (P = 0.13; Supplementary Table S2) but as expected differed in ADG (P < 0.001). Additionally, DMI (P = 0.002) and G:F (P < 0.001) differed between the groups, but RFI did not (P = 0.55). The high-ADG group had a 25.5% greater (P ≤ 0.010; Table 3) glucose nadir during the IVGTT, on average, than the medium and low-ADG groups. The effect of ADG classification on insulin time to peak (overall P = 0.036) was the result of the high-ADG group requiring a longer time to reach peak insulin concentrations than the medium ADG group (P = 0.040). The medium and high-ADG classifications were not different (P = 0.25) from each other in HCW, but both were greater (P ≤ 0.024) than the low-ADG classification.

Classification of steers based on RFI resulted in differences in DMI, G:F, and RFI (P < 0.001; Supplementary Table S2) but not initial BW (P = 0.86) or ADG (P = 0.96). Glucose AUC tended to be greater in the low RFI group (P = 0.09; Table 4) which is partially due to a greater (P = 0.040) peak glucose concentration between the low and high RFI groups (P = 0.034)

Table 4.

Effect of classification based on RFI or G:F on glucose and insulin response to an intravenous glucose tolerance test performed before the feed intake and growth measurement period and carcass characteristics

	Classification1
Item	Low	Medium	High	SEM2	P-value
RFI classification
n	17	20	17
Baseline glucose, mg/dL	108	109	103	4.8	0.60
Glucose AUC3, mg/dL × min	23,158	21,865	20,614	815.0	0.09
Peak glucose, mg/dL	313.2a	292.3ab	280.9b	7.96	0.040
Glucose clearance rate, min−1	0.028	0.027	0.027	0.0019	0.81
Glucose nadir, mg/dL	113.9	107.3	103.0	7.47	0.58
Glucose half-life, min	26.27	28.45	27.35	2.158	0.75
Baseline insulin, ng/mL	0.74	0.65	0.72	0.078	0.66
Insulin AUC3, ng/mL × min	190.8	182.1	186.7	16.90	0.93
Insulin peak, ng/mL	2.93	2.64	2.78	0.238	0.65
Insulin time to peak, min	13.67	12.75	12.19	0.983	0.56
QUICKI3	0.55	0.56	0.56	0.018	0.71
HOMA-IR3	79.0	70.0	75.3	8.84	0.75
Hot carcass weight, kg	416	421	422	7.8	0.83
12th rib fat thickness, cm	1.67	1.81	1.79	0.118	0.64
Ribeye area, cm2	93.4	88.2	90.4	1.86	0.10
USDA yield grade4	3.16	3.57	3.44	0.155	0.12
Marbling score5	603	613	572	30.8	0.53
G:F classification
n	14	27	13	—	—
Baseline glucose, mg/dL	102	105	117	5.5	0.11
Glucose AUC3, mg/dL × min	20,681b	21,321b	24,275a	866.8	0.008
Peak glucose, mg/dL	283b	290ab	319a	9.9	0.026
Glucose clearance rate, min−1	0.027	0.026	0.031	0.0021	0.14
Glucose nadir, mg/dL	101b	99b	134a	7.3	< 0.001
Glucose half-life, min	28.33	28.86	23.58	2.338	0.17
Baseline insulin, ng/mL	0.69	0.70	0.71	0.090	0.98
Insulin AUC3, ng/mL × min	173	188	196	18.8	0.65
Insulin peak, ng/mL	2.60	2.76	2.95	0.266	0.64
Insulin time to peak, min	10.77b	12.88ab	15.00a	1.021	0.017
QUICKI3	0.58	0.55	0.54	0.020	0.41
HOMA-IR3	73.0	72.8	80.3	10.22	0.81
Hot carcass weight, kg	417	417	429	8.0	0.44
12th rib fat thickness, cm	1.77	1.82	1.65	0.123	0.53
Ribeye area, cm2	91.7ab	87.5b	94.6a	1.82	0.007
USDA yield grade4	3.32	3.59	3.16	0.160	0.07
Marbling score5	610	598	581	32.6	0.82

¹Steers ± 0.5 SD of the mean for RFI or G:F were classified as medium. Those with RFI or G:F > 0.5 SD from the mean were high and those < 0.5 SD from the mean were low.

²Standard error of the mean. The most conservative (i.e., largest) estimate of SEM is reported for each item.

³AUC, area under the curve; QUICKI, quantitative insulin sensitivity check index; HOMA-IR, homeostasis model of insulin resistance.

⁴Calculated according to USDA (1997).

⁵Small⁰⁰ = 400; Modest⁰⁰ = 500; Moderate⁰⁰ = 600.

^abcSuperscript letters mean differ (P ≤ 0.05) for main effect of classification.

RFI, residual feed intake;s G:F, gain:feed.

G:F classification did not influence initial BW (P = 0.53; Supplementary Table S2) or DMI (P = 0.90) but did effect ADG (P < 0.001) G:F (P < 0.001) and RFI (P = 0.005). The high G:F steers had 17.4% greater (P = 0.008; Table 4) glucose AUC than the low G:F steers, while the medium G:F steers were intermediate and not different (P ≥ 0.10) from the other classifications. Similar to the trend in the RFI classification, the most efficient group (i.e., high G:F) had a greater peak glucose concentration than the low G:F group (P = 0.026). The low and medium G:F classifications did not differ (P = 0.98) in glucose nadir during the IVGTT, but both had lower (P ≤ 0.007) glucose nadir than the high G:F group by 34.1%, on average. The high G:F steers had a longer (P = 0.012) time to reach peak insulin concentration than the low G:F steers, with the medium G:F steers being intermediate and not different (P ≥ 0.19) from the other groups. The medium G:F steers had smaller (P = 0.004) ribeye areas than the high G:F steers, and the low group was intermediate in ribeye area and not different (P ≥ 0.20) from the other groups.

Discussion

The objective of this experiment was to determine the association of glucose metabolism and insulin resistance measured using an IVGTT with production and efficiency traits in finishing beef cattle. The measurement of glucose metabolism and insulin resistance prior to a feeding period was chosen to determine the association of the metabolic status of the animal early in the finishing phase with performance. It is unclear if the metabolic status of animals, as related to glucose metabolism and insulin resistance, early in the finishing phase is representative of their metabolic status throughout and at the end of the finishing period. Several IVGTT variables, including glucose clearance rate, insulin peak concentration, and time to insulin peak have been reported to not change over 160 d on feed while other variables, specifically fasting insulin and insulin AUC increased linearly (Joy et al., 2017). Another report measuring IVGTT variables on days 21 and 98 on feed showed most variables effected by time where some increased and others decreased (Kneeskern et al., 2016). Over a shorter period (14 d apart), there was no effect of time on IVGTT variables in cattle fed a finishing diet (Rathert-Williams et al., 2021). However, it is unclear from these data if the ranking of cattle’s IVGTT variables are altered over time. Therefore, the data presented here are interpreted as indicating that the metabolic status of cattle early in the finishing period is associated with G:F and RFI.

The use of the IVGTT allows for the evaluation of insulin resistance by measuring multiple variables that can provide evidence of alterations in glucose metabolism that are indicative of insulin resistance. The use of the IVGTT provides some advantages over surrogate indices such as QUICKI and HOMA-IR. The main advantage is related to the fact that the surrogate indices utilize only a single blood sample from fasting animals, which is a state rarely achieved in beef cattle. Additionally, these indices have been reported to not be correlated to measures from hyperinsulemic euglycemic clamp measurements (De Koster et al., 2016). The association of IVGTT variables with G:F (i.e., glucose AUC, peak glucose concentration, glucose nadir, and insulin time to peak) and RFI (i.e., glucose AUC and peak glucose concentration) reported here are interpreted to indicate that there is an association greater efficiency of feed utilization in cattle displaying characteristics of insulin resistance.

Associations of DMI with IVGTT variables

DMI was not correlated with any of the results from the IVGTT, nor associated with the IVGTT variables when classed by DMI ranking. Additionally, DMI was not correlated with the plasma glucose or insulin concentrations measured during the feed intake portion of the study. Insulin is generally regarded as a satiety signal (Wilcox, 2005) and is thought to partially potentiate the satiety signal induced by propionate (Grovum, 1995). The lack of clear associations of IVGTT parameters with DMI should not be interpreted as an indication that insulin has no role in appetite regulation in ruminants. The interrelationship between metabolic signals for appetite regulation (i.e., propionate), glucose metabolism, and insulin release cannot easily be unraveled and necessitates that glucose metabolism and insulin release be considered when explaining DMI regulation. The interrelationship of these factors is highlighted by the role of propionate in signaling satiety in ruminants (Oba and Allen, 2003; Rathert-Williams et al., 2021, 2023), while also serving as the predominant precursor for gluconeogenesis (Aschenbach et al., 2010), inducing insulin release (Manns et al., 1967), and potentially inducing insulin resistance (Rathert-Williams et al., 2021). These results from the current experiment suggest that DMI during this 63-d feeding trial was not associated with the steers IVGTT responses assessed prior to the feeding trial, which suggests that the status of insulin resistance early in the finishing phase may not predict DMI over a longer period of time. However, additional experiments are warranted to confirm these results and to investigate the role of insulin in intake regulation of finishing steers.

Associations with ADG

The current experiment demonstrated that ADG may be associated with the time to achieve peak plasma insulin concentrations following an IVGTT, as these two variables were correlated and the high-ADG group had a longer time to insulin peak in the classification analysis. Furthermore, steers with high ADG had 26% greater glucose nadir than the low and medium-ADG steers, on average. These two variables taken together indicate that steers with greater ADG take longer to release insulin from pancreatic β-cells and do not return to baseline glucose concentrations within the timeframe of the measurements. The delayed release of insulin is indicative of β-cell dysfunction which is commonly found in prediabetic animals (Petersen and Shulman, 2018). Additionally, the increased nadir in high-ADG could be due to the slower insulin response or it could indicate a decrease in insulin action resulting in blood glucose remaining elevated longer, which is a component of insulin resistance. These results lead to the postulation that cattle with increased ADG might be more insulin resistant. Insulin resistance in humans is associated with body composition changes (Carey et al., 1996); therefore, the association of growth and insulin resistance in cattle could be related to composition of the gain (i.e., fat vs. muscle accretion). It has been shown that as steers become heavier and increase fat deposition, their liver and peripheral tissues are less responsive to insulin (Eisemann et al., 1997). Additionally, adipose tissue is highly involved in the regulation of insulin resistance. In humans with dysregulated adipose, particularly in obese individuals, there is an increase in pro-inflammatory cytokines and an increase in insulin-antagonistic molecules resulting in an increase in insulin resistance (Smith and Kahn, 2016).

Previous research exploring the molecular contributions to feed efficiency have given some insight that cattle with increased growth potential could have some markers of insulin resistance. Foote et al. (2017) categorized and selected steers that had high or low ADG but displayed average DMI. In analyzing the transcriptome of the jejunum, APOB expression, which encodes apolipoprotein B, was increased in high-ADG steers compared to low-ADG. Insulin has many effects on apolipoprotein B and is associated with insulin resistance and type-2 diabetes in humans (Haas et al., 2013). Additionally, the pathway analysis indicated a differential expression of the Kyoto Encyclopedia of Genes and Genomes insulin resistance pathway (Foote et al., 2017)

Previous research that has implemented dietary treatments that altered ADG have shown no effect on cattle insulin resistance or glucose metabolism. For example, Sternbauer and Luthman (2002) fed heifers a low or high-energy diet to achieve divergent ADG. They reported no effect of dietary treatment on peak glucose or insulin, nor any difference in the insulin sensitivity indicator. Accordingly, Sternbauer and Luthman (2002) concluded that the amount of energy consumed by the heifers was too low to induce insulin resistance, which could explain the difference in the results between their experiment and the current report. Another experiment considered the effect of feeding a control diet or a high-lipid byproduct diet on finishing beef heifers (Joy et al., 2017). The control-fed heifers had 47.5% greater ADG than the high-lipid byproduct-fed heifers, but no differences were found between the faster growing control fed heifers and the slower-growing high-lipid byproduct-fed heifers on fasting plasma insulin concentrations and time to insulin peak. However, Joy et al. (2017) did not assess the direct associations of IVGTT responses to ADG making it difficult to determine how their experiment compares to the current experiment. In another experiment, the revised QUICKI (RQUICKI) approach was employed to determine how insulin resistance changes with days on feed (de Sousa et al., 2022). They determined that as days on feed increases, finishing cattle become more insulin resistant and have lower ADG. It should be noted that the surrogate index utilized by de Sousa et al. (2022) relies on glucose, insulin, and nonesterified fatty acid (NEFA) concentrations and is meant to be measured in fasting animals, yet it is unlikely the animals were fasted. Therefore, their results indicating that insulin resistance increases over time on feed is likely associated with increased concentrations of glucose, insulin, and NEFA. Even though the trend of increasing insulin resistance and decreasing ADG appears to contradict our results at first glance, one cannot make this conclusion. The decreased ADG with increasing days on feed is expected due to decreasing protein accretion and increasing fat accretion (National Academies of Sciences, 2016) and might not be directly related to insulin resistance. Furthermore, similar to Joy et al. (2017) and de Sousa et al. (2022) did not report direct associations of insulin resistance to ADG, making it difficult to compare their findings to the current study. The results of the current experiment suggest some associations with ADG and the cattle’s responses to IVGTT which agrees with previous gene expression data discussed above. However, further research is needed to determine if tissue-specific insulin resistance contributes to improved ADG of beef steers.

Associations with feed efficiency

G:F was associated with more IVGTT results than the other production traits assessed in the current experiment. G:F was positively correlated with glucose AUC, peak glucose concentration, glucose nadir, and time to reach peak insulin concentrations. Furthermore, these same traits differed by G:F classification. Additionally, RFI was negatively correlated with glucose AUC and peak glucose concentration and these results were apparent in the classification analysis as well. The trend for more efficient steers (high G:F and low RFI) to have increased glucose AUC and peak glucose concentrations, indicates that there has been a disruption in insulin action resulting in compromised glucose clearance, which is an indicator of insulin resistance. Joy et al. (2017) and de Sousa et al. (2022) both reported that G:F decreases and insulin resistance increases as days on feed increase. As discussed above for the association of insulin resistance with ADG, the association of decreased G:F with increasing insulin resistance might not be due to direct effects of insulin resistance. Based on the consistent findings of experiments like those reported by Joy et al. (2017) and de Sousa et al. (2022), one should expect an increase in insulin resistance with greater days on feed and as cattle achieve greater fat content. However, given there is no direct assessment of individual animal insulin resistance and G:F, making direct comparisons between these experiments is difficult.

In addition to determining the consistency of IVGTT variables over time as related to feed utilization, further investigation on insulin resistance and its relationship with growth performance is warranted. In humans and human models, there is evidence that insulin resistance is observed 10 to 20 yr before clinical symptoms of blood glucose dysregulation are detected (Warram et al., 1990). There has been limited specific exploration of expression of genes in the insulin signaling pathway in cattle differing in efficiency of feed utilization. In bulls, approximately 14 mo of age, research indicates that there is no difference in muscle gene expression of genes involved with insulin signaling in divergent RFI groups (Fitzsimons et al., 2014), yet there were several correlations between muscle gene expression and DMI and ADG. To the best of our knowledge, no research has been conducted to evaluate molecular changes in adipose tissue regarding insulin signaling pathways. As discussed previously, adipose tissue plays a major role in regulating insulin resistance; therefore, evaluating molecular effects on adipose tissue in relation to insulin signaling and resistance may provide insight into molecular drivers of insulin response in cattle.

The current experiment demonstrated a positive association between G:F and responses to an IVGTT that would indicate a trend toward insulin resistance. It should be noted that there are no consistent criteria to diagnose beef cattle as either insulin-resistant or not-resistant. The data presented here suggests that insulin resistance may be beneficial to feed efficiency early in the feeding period. However, de Sousa et al. (2022) reported a correlation of inflammatory markers with increasing insulin resistance. The increase in inflammation could be associated with the increased accretion of adipose tissue, which has been associated with inflammation and insulin resistance in humans (Park et al., 2014). Increased inflammatory cytokines, specifically tumor necrosis factor-α can lead to the same downregulatory effects on insulin receptor substrate-1 as insulin, but through a different mechanism (Rui et al., 2001). Therefore, the cause of insulin resistance could be different for different animals and could result in different tissues becoming insulin resistant (Piantoni et al., 2015; Mathews et al., 2016), which could alter the cost-benefit ratio of insulin resistance. It is tenuous to determine the implications of the relationship between time on feed, insulin resistance, and inflammation without further understanding of the causative agent (i.e., insulin resistance or inflammation) and the specific effects on specific tissues.

Conclusion

To the best of our knowledge, this is the first investigation into the association of glucose metabolism and insulin resistance with production traits in finishing cattle. The key finding of this investigation is that cattle with improved G:F and RFI had increased glucose AUC and peak glucose concentrations during an IVGTT and G:F was also associated with time to insulin peak. Collectively, this indicates that cattle with improved efficiency of nutrient utilization may experience greater insulin resistance than less efficient cattle. Furthermore, some associations with ADG and indicators of insulin resistance were observed, namely a positive correlation with ADG and time to reach peak insulin concentrations following an IVGTT and with high-ADG cattle having the greatest glucose nadir relative to cattle with low and medium ADG. Ultimately, these findings provide some novel insights into underlying metabolic processes that may regulate animal growth and efficiency of feed utilization.

Glossary

Abbreviations

ADG

average daily gain

AUC

area under the curve

BW

body weight

DM

dry matter

DMI

dry matter intake

G:F

gain:feed ratio

HCW

hot carcass weight

IVGTT

intravenous glucose tolerance test

NEFA

nonesterified fatty acids

NEm

net energy for maintenance

OSU

Oklahoma State University

RFI

residual feed intake

SEM

standard error of the mean

Conflict of Interest Statement

The authors declare no conflicts of interest.