Comparison of serum cortisol concentrations as indicators of the perioperative stress response in cows undergoing correction of left displaced abomasum by right flank laparotomy with omentopexy versus left flank laparoscopy with abomasopexy

Abstract

Serum cortisol is widely used as an indicator of hypothalamic–pituitary–adrenal (HPA) axis activation in cattle, although it is not specific to pain and may also reflect handling, restraint, fear, and environmental change. This study compared perioperative serum cortisol responses in dairy cows with left displaced abomasum (LDA) undergoing right flank laparotomy with omentopexy or left flank laparoscopy with abomasopexy. Forty lactating Holstein–Friesian cows were alternately allocated to laparotomy (n = 20) or laparoscopy (n = 20). Blood samples were collected immediately before surgery (T0), immediately after surgery (T1), and three hours postoperatively (T2). The linear mixed-effects model showed a significant effect of time (P < 0.001) and a significant surgical technique × time interaction (P = 0.006), whereas the main effect of surgical technique was not significant (P = 0.851). Estimated marginal means increased from 0.0505 µmol/L at T0 to 0.1649 µmol/L at T1 in the laparotomy group and from 0.0974 to 0.1312 µmol/L in the laparoscopy group, with similar values at T2 (0.1258 and 0.1247 µmol/L, respectively). Exploratory Mann–Whitney analyses showed greater T0–T1 and T0–T2 changes in the laparotomy group (P = 0.002 and P = 0.034), but no difference in T1–T2 change (P = 0.394). The temporal cortisol response therefore differed between groups, particularly immediately postoperatively. However, non-random allocation, baseline cortisol imbalance, and non-standardised preoperative handling limit causal attribution to surgical technique. Serum cortisol should therefore be interpreted as an objective marker of perioperative neuroendocrine stress rather than a specific measure of surgical pain.

Introduction

Left displaced abomasum (LDA) is a common disorder in dairy cattle and is frequently treated surgically. Non-surgical methods used to correct LDA have been associated with lower success and higher relapse rates (Wittek et al., 2009). The reported incidence of LDA in the Entre Douro e Minho Region (REDM) is 5.3% (Pires, 2004). The REDM is the main milk-producing region in mainland Portugal, accounting for approximately 47% of the country’s milk production. LDA can cause substantial economic losses through reduced milk yield during the disease and postoperative periods and through treatment costs (Fiore et al., 2018). It has also been associated with a prolonged interval from calving to first service (Jorritsma et al., 2008).

Pain is an essential protective mechanism that enables animals to detect harmful stimuli, respond to injury, and promote survival. It is a complex and multidimensional process involving sensory, physiological, neuroendocrine, and behavioural responses, with similar consequences across mammalian species (Anderson & Muir, 2005). Clinically, pain in animals is associated with actual tissue damage and induces physiological and behavioural changes indicative of stress (Broom, 2000). Nociception refers to the neural detection and processing of potentially harmful stimuli and should be distinguished from the broader experience of pain. Pain may be adaptive, promoting protection and healing, or maladaptive, when pathological changes persist after the initial injury has resolved (Muir & Woolf, 2001; Woolf, 2004). Understanding these mechanisms is essential for the prevention and treatment of pain, particularly in surgical patients, where pre-emptive analgesia can reduce the development of nociceptive responses (Anderson & Muir, 2005; Carstens & Moberg, 2000; Otto, 2001; Scholz & Woolf, 2002). In cows, the recognition of pain is particularly challenging because, as prey animals, they often show subtle signs of discomfort. Pain may be expressed through reduced feed intake, social withdrawal, altered posture, reluctance to move, changes in facial expression, increased heart and respiratory rates, and elevated serum cortisol concentrations (Guatteo et al., 2012).

Cortisol and smaller amounts of corticosterone are the most important glucocorticoid hormones secreted by the adrenal gland in many species. In general, the actions of glucocorticoids on carbohydrate, protein, and lipid metabolism result in glucose sparing and a tendency toward hyperglycaemia and increased glucose production. In addition, they decrease lipogenesis and increase lipolysis in adipose tissue, which results in the release of glycerol and free fatty acids (Edwards, 2025).

Exposure to physical or psychological stressors – such as fear, restraint, environmental changes, or perceived threats – stimulates the activation of the hypothalamic-pituitary-adrenal axis, which results in increased cortisol secretion (Herskin et al., 2004; Martin et al., 2022; Zoltick et al., 2024). Pain is also recognised as a potent activator of this endocrine pathway. Consequently, painful stimuli, such as those associated with surgical procedures, induce significant increases in cortisol concentrations (Durand et al., 2021; Sato et al., 2023). This physiological response to pain may be particularly relevant in prey species, including cattle, which exhibit few pain-related signs, thereby complicating the clinical assessment of pain and stress (Gleerup et al., 2015).

Cortisol has been studied extensively and has long been used as a physiological indicator of adrenocortical activation in response to stressors that may include fear, restraint, handling, and nociceptive procedures (Faulkner & Weary, 2000; McMeekan et al., 1998; Molony & Kent, 1997). Cortisol concentrations can increase rapidly during physical restraint, immobilisation, handling, and invasive procedures because these events may activate the stress response (Herskin et al., 2004; Radostits, 2007). In a recent systematic review of cattle castration studies, blood cortisol was the most frequently used physiological marker, and multimodal analgesia reduced the acute cortisol response (Nogues et al., 2025).

Serum cortisol is therefore widely used as a sensitive physiological marker of the neuroendocrine response to pain and stress in ruminants. However, because it lacks specificity, it should ideally be interpreted together with behavioural and clinical indicators (Coetzee, 2011; Martin et al., 2022; Stafford et al., 2002; Zoltick et al., 2024).

Serum cortisol may contribute to characterising the neuroendocrine response associated with nociceptive and other stressors, but it cannot independently quantify pain intensity. Although higher cortisol concentrations may accompany more intense painful or stressful stimuli in some experimental contexts, interpretation requires consideration of the procedure, handling conditions, behavioural responses, and other physiological variables (Bustamante et al., 2015; Fisher et al., 1996; Ley et al., 1996; Molony et al., 1993; Stilwell, 2012).

Although cortisol does not act as an analgesic, it modulates the inflammatory response and, consequently, influences pain perception (Edwards, 2025).

The administration of an α 2-adrenergic agonist (xylazine) results in lower stress and cortisol concentrations in animals (Rizk et al., 2012a; Rizk et al., 2012b; Tschoner et al., 2020).

The primary aim of this study was to compare the perioperative serum cortisol response associated with two surgical approaches for correction of LDA: one-step laparoscopic abomasopexy (Christiansen, 2004) and right flank laparotomy with omentopexy (Hannover method). We hypothesised that the laparoscopic approach would be associated with a smaller perioperative increase in serum cortisol than right flank laparotomy. Because cortisol is a non-specific marker of hypothalamic–pituitary–adrenal axis activation, the study was intended to characterise the perioperative neuroendocrine stress response rather than to validate cortisol as a stand-alone measure of surgical pain.

Materials and methods

This study included 40 Holstein–Friesian cows diagnosed with LDA from 20 dairy farms located in the REDM, Portugal, between October 2013 and February 2014. The farms had, on average, 120 lactating cows housed in free-stall barns and milked twice daily. All farms used a total mixed ration (TMR) feeding system. The cows were assigned to two groups of 20 using an alternating, non-randomised allocation procedure: odd-numbered cases underwent right flank laparotomy with omentopexy, and even-numbered cases underwent one-step laparoscopy-guided abomasopexy. Thus, treatment allocation was not based on a random sequence.

Diagnosis was established by physical examination, specifically simultaneous percussion and auscultation, and was confirmed intraoperatively by one of the authors (Giesteira).

Omentopexy is an invasive procedure performed via right flank laparotomy and involves fixation of the omentum to the right abdominal wall. This technique enables direct access to the abomasum and is associated with a reduced risk of recurrence. For laparotomy, cows were operated on in their usual housing environment and restrained at the feed barrier without being removed from their stall (Fig. 1).

Fig. 1.

Cow restrained in the farm’s feeding headlocks for laparotomy.

A modified Dirksen right flank omentopexy was performed. After clipping and aseptic preparation, the right flank was anesthetized with 100 mL of 2% lidocaine using an inverted L-block. The abdomen was opened through a 20 cm vertical right-flank incision. The displaced abomasum was passively decompressed with a needle and rubber tube, then manually repositioned to the right ventral abdomen. The greater omentum was fixed to the peritoneum, approximately 5–6 cm caudal to the pylorus, with a non-absorbable Supramid 5 suture in a simple continuous pattern. This suture incorporated the omentum into the ventral two-thirds of the incision (Giesteira et al., 2023). Postoperatively, cows received intramuscular procaine benzylpenicillin/dihydrostreptomycin for four days and a single intramuscular dose of tolfenamic acid, after the last blood sample.

Laparoscopic abomasopexy is a minimally invasive technique that consists of transabdominal suture fixation of the abomasum under laparoscopic guidance. One-step laparoscopic abomasopexy was performed following Christiansen (2004), using surgical instruments specifically developed for this approach (Dr Fritz GmbH, Germany) (Fig. 2). The cows were restrained in a standing position (Fig. 3), and two left-sided surgical sites were clipped and aseptically prepared. Local anaesthesia was performed with 20 mL of 2% lidocaine. Two small skin incisions were made for trocar placement: a 10 mm trocar for the laparoscope (Fig. 3a) and a 13 mm trocar for the instrument portal (Fig. 3b). Pneumoperitoneum was established, and the second trocar was inserted under laparoscopic guidance. The displaced abomasum was identified laparoscopically and punctured at the greater curvature with a cannula. A toggle suture was introduced into the abomasal lumen, followed by deflation of the abomasum. For repositioning and fixation, a Christiansen’s Spieker needle was advanced between the rumen and the left abdominal wall toward the right ventral paramedian region, approximately 5–10 cm cranial to the umbilicus and to the right of the midline. Its position was guided by external palpation. The ventral abdominal wall was punctured, and a Supramid suture was passed externally and connected to the toggle suture. Retraction of the suture exteriorised the toggle thread and repositioning the abomasum into its normal position.

Fig. 2.

Surgical instruments developed by Dr. Fritz; A - Laparoscope; B - Led Light Handle; C – Trocar 10 mm; D – Trocar 13 mm; E – Toggle Suture; F – Disinflation Needle; G-H – Cannula; I-J-K-L – Spieker from Christiansen.

Fig. 3.

Restraint chute developed by Giesteira, A., 2012.

The two external suture ends were passed through a gauze bandage and tied together to secure the abomasum to the ventral abdominal wall. No postoperative treatment was administered, and the gauze bandage was removed approximately 21 days after surgery (Giesteira et al., 2023).

All cows in both surgical groups were administered xylazine intravenously at a dose of 0.02 mg/kg body weight (2% xylazine) after restraint. The start of surgery was defined as the moment when local anaesthesia was administered: on the right flank in the laparotomy group and at the two trocar insertion sites on the left flank in the laparoscopy group.

The sequence of procedures was restraint of the animal, intravenous administration of xylazine, preparation of the surgical field, and local anaesthesia at the relevant incision or trocar sites. Preoperative handling and restraint were not standardised between groups: cows undergoing laparotomy remained in their usual housing environment and were restrained at the feed barrier, whereas cows undergoing laparoscopy were moved and restrained in the dedicated chute. This difference was inherent to the clinical procedures used and is relevant when interpreting serum cortisol concentrations.

Ethics statement: All animals were client-owned dairy cows with naturally occurring left displaced abomasum requiring surgical treatment. The condition was not experimentally induced, and both surgical procedures were performed for therapeutic purposes as part of veterinary clinical practice. Under Article 2(7)(b) of Portuguese Decree-Law No 113/2013, transposing Directive 2010/63/EU, non-experimental veterinary clinical practices are excluded from its scope. Blood sampling for cortisol determination did not exceed the threshold specified in Article 2(7)(f) of the same legislation. Therefore, formal project authorization under Decree-Law No 113/2013 was not required.

Blood samples (5 mL) were collected from the median coccygeal vein using an 18-gauge needle and glass tubes without anticoagulant (Vacutainer tubes) at three time points: T0 (immediately before the surgical incision), T1 (immediately after surgery), and T2 (three hours postoperatively). Serum cortisol concentrations (µmol/L) were determined at each time point. Blood samples were allowed to clot at room temperature for 30 min, placed on ice, and transported to the Fontaínhas Veterinary Clinic, where they were centrifuged at 4000 rpm for 15 min. Serum was subsequently harvested and stored at −20 °C until analysis. Cortisol analyses were performed by a certified laboratory (Segalab, Póvoa de Varzim, Portugal).

The laboratory personnel who conducted the serum cortisol analyses were not informed of the focus of our research and did not have access to the animal data.

Serum cortisol concentrations were quantified using an Immulite 2000 Veterinary Cortisol System Analyser (Siemens Healthcare Diagnostics, Llanberis, UK), employing a solid-phase competitive chemiluminescent enzyme immunoassay (CLIA). This automated immunoassay is designed for the quantitative determination of cortisol in plasma or serum samples. In the assay, cortisol in the sample competes with enzyme-labelled cortisol for a limited number of anti-cortisol antibodies immobilised on a solid phase. Following a 30-minute incubation cycle, unbound material is removed by washing, and a chemiluminescent substrate is added. The resulting light emission is measured by the analyser and is inversely proportional to the cortisol concentration in the sample, because higher endogenous cortisol concentrations reduce binding of labelled cortisol. The Immulite 2000 system automatically calculates cortisol concentrations using a calibration curve generated from standards of known concentration. The analytical sensitivity of the assay was 0.20 µg/dL (5.5 nmol/L) (conversion factor: µg/dL × 27.59 = nmol/L). The within-run coefficients of variation ranged from 5.2% to 7.4%, while total imprecision coefficients of variation ranged from 6.8% to 9.4%. The reported analytical performance characteristics support sensitive and reproducible quantification of serum cortisol concentrations in the stored samples.

Statistical analysis

Data analysis was performed using the Statistical Package for the Social Sciences (SPSS) software (Marôco, 2014).

Descriptive statistics were used to characterise the distribution of variables and included frequencies, percentages, means, and standard deviations. Inferential analyses were conducted with consideration of the assumptions required for the corresponding statistical methods.

To characterize the animals included in the study, a descriptive analysis was performed for the following variables: number of lactations, days postpartum, and body condition score, according to the surgical technique. For each variable, minimum and maximum values, mean, standard deviation, and percentiles (P25, P50, and P75) were analysed, enabling the assessment of data distribution and variability within both surgical groups.

Considering the longitudinal design, with repeated measurements obtained from the same animals at three perioperative time points (T0, T1, and T2), serum cortisol concentrations were analysed using a linear mixed-effects model (LMM). Surgical technique (laparotomy vs. laparoscopy), assessment time (T0, T1, and T2), and the interaction between these factors were specified as fixed effects. An unstructured (UN) covariance matrix was used to account for within-animal correlation. The technique × time interaction was considered the primary test of whether cortisol trajectories differed between groups. Changes in cortisol concentrations between assessment time points (ΔT0–T1, ΔT0–T2, and ΔT1–T2) were also calculated and compared between surgical groups using the Mann–Whitney U test for independent samples. These change-score comparisons were retained as supplementary exploratory analyses and were not used as substitutes for model-based longitudinal inference. Statistical significance was set at P < 0.05.

Boxplots were generated to describe the distribution of serum cortisol concentrations (µmol/L) across T0, T1, and T2 for each surgical group. They were used for descriptive visual assessment of central tendency, dispersion, interquartile range, and outlying values and were not used as a substitute for inference from the longitudinal mixed-effects model.

Results

Between October 2013 and February 2014, 40 cows with LDA underwent surgical correction by either laparotomy (n = 20) or laparoscopy (n = 20). All cows included in the study resumed milk production following surgery.

The average surgical time was 45 ± 5.1 min for the laparotomy group and 20 ± 4.5 min for the laparoscopy group. In both groups, the time between xylazine administration and the first skin incision was 15 ± 1.5 min.

In the laparotomy group, the number of lactations ranged from a minimum of 1 to a maximum of 4, with a mean value of 2.4. The 25th percentile (P25) was 1 lactation, the median (P50) was 3 lactations, and the 75th percentile (P75) was 3 lactations. Days postpartum ranged from a minimum of 2 days to a maximum of 180 days. The mean value observed was 34 days. This demonstrated substantial temporal variability among the animals evaluated. P25 was 8 days, the median (P50) was 15 days, and P75 was 27.8 days. Body condition score ranged from a minimum value of 2 to a maximum value of 4. The mean observed value was 2.9. P25 was 2.5, the median (P50) was 3, and P75 was also 3 (Table 1).

Table 1.

Number of lactations, days postpartum, and body condition score according to the surgical technique (laparotomy vs. laparoscopy).

	Laparotomy
	n	mean	sd	min	max	P25	P50	P75
Number of Lactations	20	2.4	1.1	1.0	4.0	1.0	3.0	3.0
Days Postpartum	20	34.0	49.0	2.0	180.0	8.0	15.0	27.8
Body Condition	20	2.9	0.5	2.0	4.0	2.5	3.0	3.0
	Laparoscopy
	n	mean	sd	min	max	P25	P50	P75
Number of Lactations	20	2.7	1.2	1.0	5.0	2.0	3.0	3.0
Days Postpartum	20	15.3	10.0	1.0	35.0	8.0	14.5	24.5
Body Condition	20	2.9	0.2	2.5	3.0	3.0	3.0	3.0

In the laparoscopy group, the number of lactations ranged from a minimum value of 1 to a maximum value of 5. The mean observed value was 2.7 lactations. P25 was 2 lactations, the median (P50) was 3 lactations, and P75 was also 3 lactations. Days postpartum ranged from a minimum of 1 day to a maximum of 35 days. The mean value observed was 15.3 days. P25 was 8 days, the median (P50) was 14.5 days, and P75 was 24.5 days. The body condition score ranged from a minimum value of 2.5 to a maximum value of 3. The mean value was 2.9. P25 was 3, the median (P50) was 3, and P75 was also 3 (Table 1).

Serum cortisol concentrations in the laparotomy group

At T0, serum cortisol concentrations ranged from 0.0 to 0.1 µmol/L, with a mean of 0.05 µmol/L, representing the pre-incision concentration after restraint and preoperative preparation. At T1, concentrations ranged from 0.1 to 0.4 µmol/L and the mean increased to 0.16 µmol/L. At T2, concentrations ranged from 0.0 to 0.5 µmol/L, with a mean of 0.13 µmol/L, indicating a partial decrease from T1 while remaining above the T0 mean.

Serum cortisol concentrations in the laparoscopy group

At T0, serum cortisol concentrations ranged from 0.0 to 0.4 µmol/L, with a mean of 0.10 µmol/L. At T1, concentrations ranged from 0.0 to 0.3 µmol/L, with a mean of 0.13 µmol/L. At T2, concentrations ranged from 0.0 to 0.5 µmol/L, with a mean of 0.12 µmol/L, indicating a slight decrease from T1.

To further characterise between-group differences, changes in serum cortisol concentrations (ΔT0–T1, ΔT0–T2, and ΔT1–T2) were compared between the laparotomy and laparoscopy groups using the Mann–Whitney U test for independent samples as supplementary exploratory analyses.

Discussion

Different techniques have been described for the correction of LDA, such as conservative treatments (rolling the animal), closed techniques (toggle pin) and open surgical techniques (Trent, 2017). Laparoscopic abomasopexy combines the advantages of blind toggle pin techniques (fast, minimally invasive) with the ability to visualize abdominal cavity. This leads to a lower risk of incorrect fixation (Janowitz, 1998; Roy & Rousseau, 2017).

Compared with laparotomy, one-step laparoscopy-guided abomasopexy has been associated with shorter surgical and recovery times (Janowitz, 1998; Seeger et al., 2006) and may reduce postoperative treatment requirements and costs. In the present protocol, no postoperative analgesics or antibiotics were administered to cows in the laparoscopy group, whereas tolfenamic acid was administered to the laparotomy group only after the final blood sample at T2. Therefore, tolfenamic acid did not influence the cortisol measurements reported here. Nevertheless, the absence of postoperative analgesic treatment in the laparoscopy group should not be interpreted as evidence that minimally invasive surgery is pain-free or that analgesia is unnecessary. The laparoscopic technique is feasible only when LDA is confirmed at laparoscopy. Although the left lateral abdominal cavity can be visualised, inspection of the abomasum is limited. Abomasal fistulation and local or generalised peritonitis may occur when the abomasum is not appropriately apposed to the ventral abdominal wall, and the technique may be unsuitable when adhesions are present (Seeger et al., 2006).

Omentopexy via right flank laparotomy is one of the most commonly used surgical techniques for resolving LDA, as it enables the abdominal cavity to be closely inspected (Jorritsma et al., 2008).

Pain is a potent inducer of stress and can lead to suffering (Clark et al., 1997). Stress triggers an adaptive pattern involving behavioural, neurological, endocrine, immune, haematological, and metabolic changes aimed at restoring homeostasis. The stress response prepares the animal for emergency reactions and enhances survival during an immediate threat (Anderson & Muir, 2005). Acute pain may elicit a pronounced stress response that activates the sympathetic nervous system, stimulates glucocorticoid secretion (mainly cortisol), increases basal metabolism, promotes sodium and water retention, and alters carbohydrate and protein metabolism (Carr & Goudas, 1999; Weissman, 1999).

In response to an acute stressor such as pain, homeostasis is maintained through negative feedback mechanisms that operate at multiple levels, including the amygdala, the sympathetic–adrenal axis, and the hypothalamic–pituitary–adrenal (HPA) axis. This regulation increases catecholamine and glucocorticoid levels. It also increases alertness, threat assessment, cardiorespiratory function, and cognitive performance in order to cope with immediate danger. High corticotropin-releasing hormone (CRH), serum cortisol, and vasopressin levels correlate directly with stressful or painful events and contribute to restoring physiological homeostasis (Anderson & Muir, 2005).

The relationship between cortisol levels, pain and stress in animals is well recognised and documented within the fields of animal physiology and veterinary endocrinology. Cortisol, a glucocorticoid hormone synthesised and secreted by the adrenal cortex, is a key component of the hypothalamic-pituitary-adrenal (HPA) axis and plays a central role in the neuroendocrine response to stress (Coetzee, 2011; Martin et al., 2022).

Serum cortisol measurement can provide an objective index of hypothalamic–pituitary–adrenal axis activation in cattle undergoing surgical correction of LDA, but it does not, by itself, establish the presence or intensity of pain. Previous studies indicate that local anaesthesia alone may not abolish perioperative cortisol increases, whereas combining local anaesthesia with a non-steroidal anti-inflammatory drug can attenuate the cortisol response (Erol et al., 2020; Mudroň et al., 2005; Sato et al., 2023; Tschoner et al., 2020).

Serum cortisol concentrations commonly increase during and shortly after surgical procedures as a consequence of hypothalamic–pituitary–adrenal axis activation and the associated neuroendocrine stress response. The magnitude and temporal profile of this response may be influenced by the extent of tissue injury, duration and type of surgical intervention, perioperative handling and restraint, the anaesthetic and analgesic protocol employed, and the animal’s underlying clinical condition (Coetzee, 2011; Martin et al., 2022; Mormède et al., 2007; Tschoner et al., 2020). However, cortisol is an indicator of adrenocortical activation rather than a pain-specific biomarker. This is because increased concentrations may also occur following non-nociceptive stressors such as transportation, exposure to a new environment, handling, physical restraint and social isolation (Boissy & Le Neindre, 1997; Herskin et al., 2004; Möstl et al., 2002). Consequently, cortisol measurements should be interpreted in conjunction with procedure and species appropriate behavioural and physiological indicators. These include posture, locomotion, facial expression, vocalisation, feeding and rumination behaviour, heart rate and respiratory rate. Reliable assessment of pain and perioperative stress should therefore integrate endocrine responses, behavioural alterations, physiological variables and the clinical context rather than relying on any single outcome measure (Coetzee, 2011; Martin et al., 2022; Tschoner et al., 2020).

Consistent with this multimodal approach, the UNESP-Botucatu Unidimensional Pain Scale was developed and validated for assessing acute postoperative pain in cattle following orchiectomy. The scale evaluates five behavioural domains: locomotion, interactive behaviour, activity, appetite and miscellaneous pain-related behaviours. Pain scores increased after surgery and decreased following rescue analgesia, confirming the responsiveness of the scale. The instrument also demonstrated strong validity, excellent internal consistency and moderate-to-very good observer reliability. A score of ≥5 on the 0–10 scale was proposed as the threshold for rescue analgesia. Although the scale was validated in cattle undergoing orchiectomy rather than LDA correction, it illustrates the value of structured behavioural assessment as a complement to cortisol measurement when evaluating perioperative pain in cattle (de Oliveira et al., 2014).

Xylazine, an α₂-adrenergic agonist, reduces central norepinephrine release and can decrease stress-related cortisol concentrations while producing sedation and muscle relaxation (Rizk et al., 2012a; Rizk et al., 2012b). In cattle, sedation begins approximately 3 min after intravenous administration, and the serum half-life is approximately 30–36 minminutes (Ebert et al., 2002; Löscher, 2006). However, sedation and reduced responsiveness should not be interpreted as evidence of adequate analgesia. Calves dehorned under xylazine alone showed little behavioural resistance but still developed marked cortisol responses, indicating persistent nociceptive stress (Stilwell, 2006).

The clinical use of α₂-adrenergic agonists is limited by their relatively short analgesic duration and potential adverse effects. These include bradycardia, hypotension, ruminal atony, increased diuresis, inhibition of insulin release, and uterine contractions. In the present protocol, all cows received xylazine at 0.02 mg/kg intravenously together with local anaesthesia at the surgical incision sites. Tolfenamic acid was administered only to the laparotomy group, three hours after surgery and after the final blood collection.

Preoperative administration of low-dose xylazine has been associated with lower intraoperative serum cortisol concentrations during two-step laparoscopic correction of displaced abomasum (Tschoner et al., 2020). Xylazine administered at 0.02 mg/kg intravenously may therefore be included as part of a multimodal perioperative protocol combining local anaesthesia and a non-steroidal anti-inflammatory drug. However, the recommendation should be limited to the surgical techniques for which supporting evidence is available, rather than applied universally to all abomasal procedures (Tschoner et al., 2020).

Previous studies have reported mean preoperative cortisol concentrations of approximately 0.05 µmol/L in cows with left displaced abomasum 90 min before surgery (Tschoner et al., 2020). In the present study, the T0 concentration in the laparotomy group was close to this value, whereas the mean T0 concentration in the laparoscopy group was approximately twice as high (Table 3). This pre-incision imbalance is important for interpretation. Cows in the laparoscopy group were moved from their usual environment and restrained in a dedicated chute, whereas cows undergoing laparotomy remained in the barn and were restrained at the feed barrier. Because environmental change, handling, and restraint can independently activate the HPA axis, these procedural differences may have contributed to the higher T0 cortisol concentration in the laparoscopy group. However, the present design cannot determine the relative contribution of handling, restraint, clinical condition, and surgical approach; the T0 imbalance should therefore be regarded as possible source of confounding rather than attributed definitively to the laparoscopic procedure.

Table 3.

Estimated marginal means of serum cortisol concentrations (µmol/L) according to surgical technique (Laparotomy vs. Laparoscopy) and assessment time (T0, T1, and T2).

Surgical technique	T0	T1	T2
Laparotomy	0.0505	0.1649	0.1258
Laparoscopy	0.0974	0.1312	0.1247

Note. Values are presented as estimated marginal means (µmol/L).

The LMM revealed no significant main effect of surgical technique on serum cortisol concentrations (F(1. 38) = 0.036; P = 0.851). This indicates that there was no statistically significant overall difference between groups when all assessment times were considered together; it does not demonstrate equivalence between the procedures or eliminate the potential influence of baseline and handling-related differences.

Assessment time had a significant effect on serum cortisol concentrations (F(2. 38) = 20.617; P < 0.001), demonstrating marked perioperative changes over time. A significant surgical technique × time interaction was also identified (F(2. 38) = 5.773; P = 0.006), indicating that the temporal cortisol profile differed between groups. Given the non-randomised allocation and the different preoperative handling and restraint conditions, this interaction should be interpreted as an association between treatment-group context and cortisol trajectory rather than as an isolated causal effect of surgical technique.

The estimated marginal means presented in Table 2 illustrate the temporal changes in mean cortisol concentrations according to surgical technique and assessment time. This supports the significant interaction identified by the linear mixed-effects model.

Table 2.

Type III tests of fixed effects from the linear mixed-effects model for serum cortisol concentrations (µmol/L).

Effect	df	F	p
Surgical technique	1. 38	0.036	0.851
Time	2.38	20.617	<0.001
Surgical technique × Time	2.38	5.773	0.006

Note. F = F statistic; df = degrees of freedom. Significant effects are indicated by bold p-values (p < 0.05).

Estimated marginal means demonstrated an increase in serum cortisol concentrations immediately after surgery (T1) in both surgical groups, followed by a decrease at T2. In the laparotomy group, the estimated mean serum cortisol concentration increased from 0.0505 µmol/L at T0 to 0.1649 µmol/L at T1 and subsequently decreased to 0.1258 µmol/L at T2. Similarly, in the laparoscopy group, serum cortisol concentrations increased from 0.0974 µmol/L at T0 to 0.1312 µmol/L at T1, followed by a slight decrease to 0.1247 µmol/L at T2.

No significant overall main effect of surgical technique was observed. However, the significant surgical technique × time interaction (P = 0.006) indicates that the temporal pattern of serum cortisol concentrations differed between groups. Estimated marginal means were already higher at T0 in the laparoscopy group (0.0974 µmol/L) than in the laparotomy group (0.0505 µmol/L). From T0 to T1, the increase was more pronounced in the laparotomy group, whereas by T2 the estimated means were nearly identical (0.1258 and 0.1247 µmol/L, respectively). Accordingly, the interaction supports different temporal profiles but does not establish that surgical technique alone caused the difference because allocation and preoperative handling/restraint were not fully comparable.

In the supplementary exploratory change-score analyses, the increase in serum cortisol from T0 to T1 was greater in the laparotomy group than in the laparoscopy group (U = 86.5, Z = −3.070, P = 0.002) (Table 4). The T0–T2 change also differed between groups (U = 121.5, Z = −2.124, P = 0.034), whereas the T1–T2 change did not (U = 168.5, Z = −0.853, P = 0.394). Because change-score analyses do not resolve the non-randomised allocation, T0 imbalance, or handling/restraint confounding, these findings should be considered supportive exploratory analyses of the temporal pattern identified by the LMM rather than independent evidence of a causal advantage of one technique.

Table 4.

Comparison of changes in serum cortisol concentrations (Δ) between the Laparotomy and Laparoscopy groups using the Mann–Whitney U test.

Interval	Laparotomy (Mean rank)	Laparoscopy (Mean rank)	U	Z	p
ΔT0–T1	26.18	14.83	86.5	−3.070	0.002
ΔT0–T2	24.43	16.58	121.5	−2.124	0.034
ΔT1–T2	18.93	22.08	168.5	−0.853	0.394

Note. ΔT0–T1 = T1 − T0; ΔT0–T2 = T2 − T0; ΔT1–T2 = T2 − T1. U = Mann–Whitney U statistic; Z = standardized test statistic. Statistically significant p-values (p < 0.05) are shown in bold.

Taken together, the LMM and exploratory change-score analyses show that serum cortisol concentrations changed over time and that their temporal pattern differed between the two groups. The data are compatible with a more pronounced immediate postoperative cortisol increase in the laparotomy group, but causal attribution of this difference to surgical technique alone is not justified.

Overall, the linear mixed-effects model and the graphical analysis showed different temporal serum cortisol profiles in the laparotomy and laparoscopy groups. The boxplots were visually consistent with a greater immediate postoperative increase and wider dispersion of cortisol values in the laparotomy group (Fig. 4). Because the boxplots are descriptive, interpretation should remain anchored in the mixed-effects model and in the methodological limitations of the study.

Fig. 4.

Graphical distribution of serum cortisol concentrations (µmol/L) at the different time points (T0, T1 and T2) according to surgical technique (laparotomy vs. laparoscopy).

Principal contribution and study limitations

The principal contribution of this study is the characterisation of repeated perioperative serum cortisol measurements in cows undergoing two clinically relevant approaches for LDA correction, analysed with a mixed-effects model appropriate for within-animal repeated measures. The study therefore contributes original clinical data on the perioperative neuroendocrine stress response associated with these surgical contexts; it does not validate serum cortisol as a stand-alone measure of surgical pain.

Several limitations restrict causal interpretation. First, treatment allocation was alternating rather than randomised. Second, pre-incision serum cortisol concentrations were higher in the laparoscopy group, and preoperative handling and restraint differed between groups; because cortisol is sensitive to environmental change, handling, and restraint, these factors may have contributed to both the T0 imbalance and the subsequent cortisol profile. Third, no complementary validated behavioural pain scale or other structured clinical pain measure was collected, so pain cannot be inferred from cortisol alone. Finally, the dataset was collected in 2013–2014 and no formal prospective sample-size calculation was performed. These limitations reduce the strength of causal attribution and should be considered when assessing the precision and generalisability of the findings.

The magnitude of surgical pain in cattle is influenced by the nature and extent of tissue injury, the surgical technique employed, the duration of the procedure, and the experience and technical competence of the practitioner (Anderson & Edmondson, 2013; Anderson & Muir, 2005). Appropriate pain prevention and management are required on ethical and animal-welfare grounds and are supported by professional, clinical, societal, and legal considerations, depending on the jurisdiction. Effective analgesia may also facilitate recovery and reduce the adverse physiological and behavioural consequences of untreated pain. However, its economic benefits may vary according to the procedure and production context (Kleinhenz et al., 2021; Steagall et al., 2021). The therapeutic objective should be to prevent or attenuate peripheral and central sensitisation, minimise the development of maladaptive or pathological pain, and restore normal function while preserving the protective role of physiological nociception (Anderson & Muir, 2005; Zoltick et al., 2024).

Conclusions

Serum cortisol provided an objective indicator of the perioperative neuroendocrine stress response in cows undergoing surgical correction of left displaced abomasum, but it should not be regarded as a specific measure of surgical pain. Cortisol concentrations changed significantly over time, and the temporal pattern differed between the two surgical groups, with a greater T0–T1 increase in the laparotomy group. However, because allocation was alternating rather than randomised, pre-incision cortisol concentrations differed between groups, and preoperative handling and restraint were not standardised, the observed cortisol differences cannot be attributed exclusively to surgical technique. These findings support an association between the surgical-treatment context and the perioperative cortisol response, while reinforcing the need to assess pain using complementary behavioural, physiological, and clinical indicators. Prospective randomised studies with standardised handling and restraint, multimodal pain assessment, and a pre-specified sample-size calculation are warranted.

CRediT authorship contribution statement

A.M. Giesteira: Writing – original draft, Validation, Methodology, Investigation, Conceptualization. M.S. Lima: Writing – review & editing, Formal analysis, Conceptualization. A.C. Maurício: Writing – review & editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.