Induction of hypocalcaemia and evaluation of reticuloruminal motility using a three-axis accelerometer

Keywords: dairy cows, detection of periparturient disease, calcium, forestomach motility in cattle, induced hypocalcaemia, remote sensing, smart farming, three-axis accelerometer.

Rumen contraction is a unique physiological motility observed in ruminants to mix feed materials and to facilitate microbial fermentation (Foster 2017). It can be evaluated in the left paralumbar fossa via auscultation and palpation, with a rate of two to four contractions per 2 min in healthy individuals. The contraction is centrally coordinated by the vagus nerve, with regular stimulations of reticuloruminal motility, and it can be reduced when there are acidic ruminal contents, pain from illness, or a stressful environment (Ivany et al. 2002; Odongo et al. 2006; Bedford et al. 2020).

Calcium is an essential factor that excites the smooth muscles in the gastrointestinal tract including the forestomach in ruminants (Armstrong and Cota 1999; Perrino 2016). Decreased concentrations of extracellular blood calcium are related to forestomach hypomotility and inappetence in ruminants (Jørgensen et al. 1998; Goff et al. 2020). Particularly in dairy cows after parturition, the concentration of blood calcium decreases within 1 day after calving due to milk production and consequent demand, and the concentration cannot return to the normal range when there is not enough capacity for reabsorption or mobilisation (Jørgensen et al. 1998; Martinez et al. 2014). In clinical hypocalcaemia, the lack of ruminal motility can induce metabolic diseases such as ketosis, and the risks of abomasal displacement and retention of placenta increase (LeBlanc et al. 2005; Seifi et al. 2011; Rodríguez et al. 2017). Even though there are no clinical signs, cows with decreased calcium concentrations are vulnerable to additional infections and show reduced performance in further reproduction (Venjakob et al. 2018).

With the recent development of information and communication technology, quantitative and continuous measurement of physiological parameters has been enabled wirelessly. Additionally, in the cattle industry, various physiological changes and behaviours can be effectively monitored using various types of biosensors (AlZahal et al. 2011; Stevenson et al. 2014; Wolfger et al. 2015; Neethirajan 2017). Biosensors have been applied to measure vital signs and physical activities such as body temperature, respiratory rate, lameness and exercise, and to evaluate productivity indicators such as ovarian cycles and delivery time. Among them, the implanted electrodes recorded the contraction patterns of the smooth muscle to analyse the myoelectrical activities of the reticulo-rumen in sheep, and they enabled wireless electromyography (EMG) telemetry registration (Wierzbicka et al. 2021). In contrast, a bolus-type accelerometer in the reticulum and a force transducer on the ruminal sac could detect both the frequency and amplitude of rumen motility accurately and effectively (Arai et al. 2019; Hamilton et al. 2019; Choi et al. 2020).

The present study aimed to monitor rumen motility using three-axis accelerometer (3XA) in a case of hypocalcaemia in dairy cows. For this purpose, hypocalcaemia has been experimentally induced by infusion of ethylenediamine–tetraacetic acid disodium salt dihydrate (Na₂EDTA), and confirmed by evaluation of the blood calcium concentration.

Six non-lactating, non-pregnant Holstein cows with a mean age of 79 ± 18 (s.e.m.) months were used in the experiments. The cows were housed together in the barn with headlock feeders and allowed to access water without limitation. The same amounts of grass hay and concentrated feed were supplied for each cow twice a day. All protocols for animal experiments were validated by the Institutional Animal Care Use Committee of Seoul National University (ethical approval number SNU-190922-3, date of approval 25 November 2019).

The bolus-type biosensor, which incorporated a 3XA and a temperature sensor (uLikeKorea Co., Inc.), was prepared (Choi et al. 2020). At least 1 month before the experiments, the bolus was inserted by oral administration and settled in the reticulum.

A 2 × 2 crossover experiment was designed, and the cows were assigned to either the treatment (TRE) or control (CON) group. For the first treatment, three cows were assigned to the TRE group, and the other three cows were assigned to the CON group. For the second treatment after 1 week, the cows belonging to the TRE and the CON switched groups and received the opposite treatment. One day before each treatment, intravenous catheters (EQUIVET Hiflow long-term IV catheter 14 G × 5.25 inch, KRUUSE) were inserted in one side of the jugular vein and fixed with skin sutures, and 1 mL of heparinised saline was flushed into the catheter to prevent blood clotting. In the TRE, 13% Na₂EDTA (Sigma-Aldrich) solution was infused at a rate of 80 mg/kg for 1 h (total of 48 g of Na₂EDTA per cow), and the same volume of physiological saline (DAI HAN PHARM) was infused into the CON through the intravenous catheter. The total amount of the administered Na₂EDTA was calculated by taking the weight of the cows as approximately 600 kg. Both groups were administered simultaneously in both of the first and the second treatments. Following infusion of the Na₂EDTA solution and physiological saline for 1 h, all cows were released from the stanchion and had access to water.

The magnitude of each axis was recorded as the X, Y and Z values (longitudinal, horizontal, and vertical axes respectively), and each acceleration was influenced by the force of gravity. The vector magnitudes (V) and the amount of changes over time (V₁ and V₂) were calculated as follows:

The maximum V values every 2 min were recorded and transferred using a wireless network. To evaluate reticular motility, the amount of change in the V values was primarily calculated as V₁. However, the magnitude of the Y-axis was changed most sensitively when the 3XA rotated along the longitudinal axis of the device (Choi et al. 2020). Unpredictable fluctuations of V₁ were observed even while reticuloruminal motility did not occur, and it was assumed that the bolus sensor itself was sometimes rotating by the flow of the rumen fluid; thus, the calculation was modified to exclude rotational movement of the biosensor, and the V₂ value was calculated to be used as the main acceleration value.

The transferred 3XA data were collected from 2 h before until 6 h after the start of the infusion. The average pre-infusion V₂ for 2 h and the averages of post-infusion V₂ at every 30 min interval for 6 h were calculated.

Cows were held in a headlock for 1 h of the infusion, and blood sampling from the coccygeal vein with a heparin vacuum tube (lithium heparin tube; BD Vacutainer) was performed right before infusion (0 h) and at 1, 3, 6 and 24 h of the infusion. The collected blood samples were centrifuged at 2000g for 15 min at a room temperature of 18°C, and plasma concentrations of albumin (Alb), total cholesterol (tChol), total calcium (tCa) and inorganic phosphate (iP) were analysed using a biochemistry analyser (BS-400; Mindray, Shenzhen, China).

The average values of accelerations and the blood analysis results were statistically analysed with two-way repeated-measures ANOVA followed by the Holm–Sidak method for multiple comparisons among groups over time. All of the statistics were performed using SigmaPlot 12.5 (Systat Software Inc., San Jose, CA, USA), and the P-value was set at 0.05.

After each treatment for 1 h infusion had been completed, there were no cows lying down, and no changes in clinical signs were observed. The concentrations of blood samples collected at 0 h ranged between 3.4 and 4.3 g/dL for Alb, 70 and 98 mg/dL for tChol, 2.0 and 2.3 mmol/L for tCa, and between 1.5 and 2.5 mmol/L for iP, and there were no significant differences between CON and TRE. Furthermore, the concentrations of Alb and tChol were not significantly different between CON and TRE at 1, 3, 6 and 24 h (Fig. 1a, b). However, in TRE at 1 h, the end of the Na₂EDTA infusion, the average concentration of tCa was significantly decreased (1.23 mmol/L) compared with TRE at 0 h (2.20 mmol/L, P < 0.001) and CON at 1 h (2.18 mmol/L, P < 0.001; Fig. 1c). The concentration of tCa was increased at 3 h (1.89 mmol/L), and significant differences were still observed compared with TRE at 0 h and CON at 3 h (P < 0.001 for both). The concentration at 6 h (2.07 mmol/L) was significantly different only from that at 0 h (P < 0.05), and, finally, it was confirmed to be recovered to the level of pre-infusion at 24 h (2.23 mmol/L). The average concentration of iP was also decreased following Na₂EDTA infusion in TRE (Fig. 1d), and there were statistically significant differences at 1 h from that in TRE at 0 h and CON at 1 h (P < 0.001 for both). The results of two-way repeated ANOVA are shown in Table 1.

Fig. 1.  The plasma concentrations of (a) albumin, (b) total cholesterol, (c) total calcium, and (d) inorganic phosphate at 0, 1, 3, 6 and 24 h after infusion. The values represent the means ± s.d. Significant differences are shown compared with the pre-infusion concentration (^†P < 0.05) and the CON at each time point (*P < 0.05).

Table 1.  Results of the two-way repeated-measures ANOVA for plasma concentration of total calcium and inorganic phosphate and V₂ values of the 3XA in the reticulum.

As the representative results of CON and TRE, the magnitudes of three axes in 2 min intervals were obtained from one cow, and the V and V₂ of the 3XA were calculated (Fig. 2a). The acceleration values of each axis showed irregular fluctuations in both the CON and TRE of one cow. Among the V₂ values, a decrease in amplitude was observed from approximately 30 min in TRE and maintained for a few hours. For the average of V₂ values in each group (Fig. 2b), the acceleration value of TRE was decreased to 200 or less after the infusion started and maintained for approximately 1 h. In that period, the V₂ values of CON and TRE were separated from each other and showed overlapping values from approximately 2 h.

Fig. 2.  The accelerometer output in 2 min intervals. The values of the three axes (X, Y and Z) were recorded, and the amount of vector (V) and its changes over time (V₂) were calculated. (a) A representative figure of one cow is shown, and (b) the mean V₂ values of six cows are represented in 2 min intervals.

In the V₂ calculated at 30 min intervals (Fig. 3), the values for 2 h before infusion were approximately 250 in both CON and TRE, but the value of TRE was decreased to less than 200 from 0.5 h to 1.5 h after the infusion. Compared with CON, significant differences were observed from 1 to 1.5 h and from 1.5 to 2 h (P < 0.05 for both). Statistical significance was obtained comparing CON and TRE over time (P < 0.05; Table 1).

Fig. 3.  The averages of V₂ values in 30 min intervals. The values represent the means ± s.d. After Na₂EDTA infusion, the V₂ values were compared between CON and TRE at each time point, and significant differences are shown from 1 to 2 h (*P < 0.05).

The reduced motility of gastrointestinal tract is one of the typical symptoms that appears in most disease states in cattle, and it is an important biological sign, like body temperature and vitality. In periparturient dairy cows, changes in rumen motility can be used as an indicator of the occurrence of related disorders. In this study, therefore, we evaluated whether biosensors located in the reticulum of dairy cows can detect decreases in reticuloruminal motility and whether wirelessly transmitted data are practically helpful.

Na₂EDTA solution has been used in numerous studies of bovine hypocalcaemia (Jørgensen et al. 1999; Mellau et al. 2001), and when a chelating agent such as citrate or EDTA is administered, it diminishes ionised plasma Ca concentrations successfully in young and adult cattle (Desmecht et al. 1995; Martinez et al. 2014; Sasaki et al. 2014; Ro et al. 2020). In this study, all of the concentrations of blood samples collected at 0 h were in a normal range, but the infusion of Na₂EDTA at a total amount of 48 g for 1 h was sufficient to induce hypocalcaemia in 600 kg cows. Assuming that the largest amount of EDTA was administered at 1 h, the end of the EDTA infusion, the decreased tCa concentration reflects the chelating action of EDTA. The decreased Ca concentration recovered gradually from 1 h after the infusion (Fig. 1c). Parathyroid hormone (PTH) is secreted in response to decreased blood Ca concentrations, and the circulating concentration of Ca is increased by resorption from bone and reabsorption from the kidney (Kronqvist et al. 2011; Blaine et al. 2015; Ro et al. 2020). The concentration of iP was also decreased after Na₂EDTA infusion (Fig. 1d), which resulted from the PTH action on renal phosphate excretion (Mellau et al. 2001; Blaine et al. 2015). During the experimental period, the concentrations of Alb and tChol in both CON and TRE were maintained within a normal range without significant changes (Fig. 1a, b). The results indicated that the decreased tCa concentration in TRE was not due to hypoalbuminemia and was also not related to other blood parameters (e.g. cholesterol catabolism) following a short treatment time of 1 h (Onifade et al. 2005; Gallo et al. 2016).

Since the 3XA indwelling in the reticulum measures the reticuloruminal motility, which reflects the vector value according to the change of position as it is, data processing was required to show the movement of the 3XA. The V₂ value was used to better indicate the difference in reticuloruminal motility than V or V₁ by lowering the baseline of the resting period (Choi et al. 2020), and in this study, the V₂ value showed different accelerations between Con and TRE (Fig. 2). Among the acceleration values of V₂ with 30 min intervals (Fig. 3), the V₂ value fell below 200 from 0.5 h to 1.5 h, similar to the results after xylazine administration in a previous study (Choi et al. 2020). Statistical significance was evident from 1.5 h, but four of six cows started to show a decrease in V₂ at 0.5 h, and five of six cows showed a large decrease at 1 h compared with 0 h. Since the influx of extracellular Ca is important in reticuloruminal contractions (Perrino 2016), the decrease can be immediately reflected. However, individual abilities to mobilise Ca in the body were not considered in the present study.

There are several limitations to this study. First, the experiment was conducted to induce hypocalcaemia in a small number of cows, and they are not representative of lactating or pregnant ones. Second, the bolus-type sensor with 3XA located in the reticulum detects and records reticuloruminal motility indirectly according to the movement of reticular walls or the flow of its contents. Recently, electrodes were surgically attached to sheep to assess the EMG activity of reticuloruminal smooth muscle directly, and the transmitted data could be recorded wirelessly for ∼2 months (Wierzbicka et al. 2021). So as to prevent a productivity decrease in cattle caused by forestomach problems, further studies to develop a non-invasive and accurate method based on these studies are necessary.

The use of acceleration data wirelessly transmitted by an intrareticular biosensor with 3XA allows to record the reduction of rumen motility caused by experimentally induced hypocalcaemia.

The data that support this study cannot be publicly shared due to ethical or privacy reasons and may be shared upon reasonable request to the corresponding author if appropriate.

The author declares no conflicts of interest.

This work was supported by IPET through the Advanced Production Technology Development Program, funded by MAFRA (318103-2).