Evaluation of ear skin temperature as a cow-side test to predict postpartum calcium status in dairy cows

Subclinical hypocalcemia is considered a gateway disease that increases susceptibility to other metabolic and infectious diseases in transition dairy cows. In the absence of a cow-side test, however, it is difficult to identify hypocalcemic cows. The objective of this study was to evaluate ear skin temperature as a diagnostic predictor of serum calcium concentration. We conducted a cross-sectional study on 7 commercial dairy farms, involving 251 cows 0 to 48 h after calving. Skin temperature of the ears (STEar) was scored manually by palpating both ears. An infrared thermometer was used to measure ear temperature, skin temperature on the coxal tuber (STCox), and ambient temperature. Rectal temperature was measured using a digital thermometer. A blood sample was drawn to determine serum calcium concentration. Hypocalcemia was defined as serum calcium below 2.0 mmol/L, irrespective of clinical symptoms. Serum calcium concentration <2.0 mmol/L in connection with clinical symptoms was defined as clinical milk fever; serum calcium concentration <2.0 mmol/L without clinical symptoms was defined as subclinical hypocalcemia. Multivariate analysis using the GENLINMIXED procedure and receiver operating characteristic analysis were performed to evaluate whether serum calcium concentration could be predicted using ear temperature and other temperature estimates. The prevalence of hypocalcemia was 3.3, 27.3, 32.8, and 69.6% for cows in first, second, third, and fourth or greater lactation, respectively. None of the cows in first and second lactation had clinical milk fever. The prevalence of clinical milk fever was 6.0 and 20.3% for cows in their third and fourth or greater lactation, respectively. A decrease in ear temperature of 0.39°C [95% confidence interval (CI): 0.25-0.54] was associated with a decrease of 0.1 mmol/L in serum calcium concentration. Ambient temperature, however, was a major confounder for ear temperature. With an increase in ambient temperature of 1°C, STEar rose by 0.78°C (95% CI: 0.67-0.90). Hypothermia was more pronounced in clinical milk fever (median 21.8°C; interquartile range 14.7-27.0°C) compared with subclinical hypocalcemia (median 27.6°C, interquartile range 22.1-30.8°C). All temperature estimates had only accurate test characteristics based on their area under the curve for prediction of subclinical hypocalcemia (area under the curve for STEar, STCox, and rectal temperature were 0.641, 0.668, and 0.606, respectively) when cows with clinical milk fever were excluded. Although ear temperature has been associated with serum calcium concentration, ear temperature cannot be recommended for diagnosis of subclinical hypocalcemia.