ETD

• cows
• ketosis
• metabolic markers
• monitoring
• sensor

The transition period (TP), spanning three weeks before and after calving, represents the most critical phase for dairy cows due to physiological and metabolic changes that lead to negative energy balance (NEB). This condition, combined with immune system dysfunction during TP, predisposes animals to metabolic disorders, with ketosis being the most significant. Currently, the diagnostic gold standard for ketosis is the measurement of blood β-hydroxybutyrate (BHB), an invasive procedure impractical for continuous herd-level monitoring. These limitations highlight the need for novel diagnostic tools. The Precision Livestock Farming (PLF) system, based on the use of sensor technologies, may enable non-invasive, continuous and real-time monitoring. Research protocol 1: The study aimed to evaluate the feasibility of microneedle (MN) technology in dairy cows by assessing skin thickness at potential application sites and to identify the optimal implantation area based on ultrasonographic, histological and practical criteria. Fifteen healthy Italian Holstein cows were selected at two slaughterhouses. Four anatomical areas potentially suitable for the implantation of MNs were identified: the pinna (regio auricularis); the parotid region (fossa retromandibularis); the groin region (regio inguinalis) and the ventral portion of the tail base (radicis caudae). The protocol was composed of: (1) an in vivo ultrasonographic study, aimed to define the best ultrasonographic execution protocol (Part 1A) and aimed to measure skin thickness (Part 1B); (2) a post-mortem histological analysis for the examination of skin samples. From the comparative analysis of four ultrasonographic protocols, the best one was based on the use of a portable ultrasound device, a hockey stick probe and a gel pad interposed between the probe and the skin (MET-2 GEL). Ultrasonographic measurements (120 in total) from 60 images revealed the following median skin thicknesses: regio auricularis 2.46 mm; fossa retromandibularis 4.85 mm; regio inguinalis 3.77 mm; radicis caudae 3.33 mm. The histological analysis (1200 measurements) of 60 skin samples confirmed the considerable variability and thickness of the tissues; the distance from the skin surface to the end of the dermis showed median values of: 6789.04 µm for radicis caudae; 7315.39 µm for regio inguinalis; 3379.02 µm for the regio auricularis and 9250.21 µm for the fossa retromandibularis. In conclusion, the skin thickness detected in all examined areas is incompatible with the length of current MNs, which typically do not exceed 1.5-2 mm. Therefore, MN technology does not currently appear to be applicable to dairy cows, suggesting the need to develop devices with deeper penetration capability. Research protocol 2: The study evaluated the application of FreeStyle Libre 3 (FL3) sensors for continuous glucose monitoring in dairy cows, comparing the readings provided by FL3 with the gold standard, glucometer, and testing their operational duration and persistence in four anatomical sites: regio auricularis; fossa retromandibularis; regio inguinalis and radicis caudae. Twelve healthy Italian Holstein cows were included. Blood samples were taken every 15 minutes for a total of 6 hours (from T0 to T24). Three hours after the start of the test (T12), a 50% glucose bolus was administered intravenously at a dose of 0.3 g/kg body weight. The results showed variable accuracy of the FL3 sensors depending on the site. The parotid region demonstrated the best agreement with the reference method (average ICC = 0.864), followed by the ear (ICC = 0.681), while the groin (ICC = 0.573) and tail base (ICC = 0.600) showed fair agreement. The mean glucose concentration detected by the FL3 ranged from 70.29 mg/dL (groin) to 94.37 mg/dL (ear), compared to 60.66 mg/dL for the gold standard. The tail base was the best site for functionality and persistence: 67% of sensors functioned for the entire 6-hour trial, and the mean post-trial retention was 1.0 ± 1.0 days. In contrast, sensors on the ear and parotid region detached within a few hours. In conclusion, although the tail base and parotid region are the most promising sites, the significantly reduced retention time compared to human use currently makes the application of this technology for continuous monitoring in dairy cows impractical. Research protocol 3: The study investigated the relationship between the fatty acid (FA) profile of milk and blood β-hydroxybutyrate (BHB) concentrations in dairy cows during the first 21 days postpartum. Fifty-three cows were included, classified based on BHB levels as healthy (H, BHB < 1.2 mmol/L; n=35) or with subclinical ketosis (SK, BHB ≥ 1.2 mmol/L; n=18). The exclusion criteria were the onset of diseases other than ketosis during the entire study period. Clinical examinations and milk samplings were performed at four different time points: T0, T7, T14, and T21 (within 3 days of calving and at 7, 14, and 21 days postpartum, respectively). The SK group showed higher median BHB concentrations at all sampling times (T0=1.30 mmol/L; T7=1.55 mmol/L; T14=1.60 mmol/L; T21=1.50 mmol/L) compared to the H group (T0=0.70 mmol/L; T7=0.80 mmol/L; T14=0.75 mmol/L; T21=0.70 mmol/L). Multivariate factor analysis identified four latent factors from the milk FA composition: Factor 1 (biohydrogenation index) showed no significant effects of time, BHB group, or their interaction; Factor 2 (de novo synthesis) was significantly influenced by ketosis, with lower scores in SK cows; Factor 3 (desaturation index) showed a significant effect of time, with scores increasing at T14 and T21 in both groups; Factor 4 (C4:0) showed a significant time × group interaction, with a peak at T7–T14 in H group and higher values at T0 in SK group.The results confirm that the milk fatty acid profile reflects the metabolic alterations of subclinical ketosis, offering a potential non-invasive monitoring tool. Research protocol 4: The study aimed: 1) to evaluate possible differences in blood-gas parameters between healthy cows and cows with diet-induced ketosis during transition period; 2) to evaluate possible differences in urinary analytes between the two groups; 3) to investigate the relation between urinary fractional electrolyte excretion and blood-gas analysis in the same population. Thirteen cows were included and divided into a control group (CG, n=7) and a ketosis induction group (KG, n=6). Clinical examinations, blood and urine samplings were performed at five different time points: T0, T7, T14, T21, and T28 (corresponding to 1, 7, 14, 21, and 28 days after calving, respectively). Despite different diet, only 67% (4/6) of cows in the KG developed subclinical ketosis (BHB ≥1.2 mmol/L), while 43% (3/7) of the CG showed similar BHB levels. Statistically significant differences between CG and KG were observed: at T0 for BCS, ionized calcium, HCT, CHGB and urinary concentrations of chloride and calcium; at T7 for BCS, HCT and CHGB; at T14 for BCS, urinary creatinine and urinary calcium concentrations; at T21 for BCS, pH, pH(T), HCT, CHGB and urinary ketone bodies; at T28 for BCS, HCT, CHGB, urinary concentrations of ketone bodies and creatinine. A strong correlation between blood BHB and urinary ketone bodies was observed in the KG (rho = 0.77). In conclusion, the combined monitoring of blood-gas parameters and urinary analytes during the TP provided complementary insights into the metabolic adaptations of dairy cows.