Vitamin A (retinol) is an essential micronutrient not only to ascertain vision but also to modulate growth and development [1–4]. Cattle as well as other herbivorous animals are unable to synthesize retinol de novo and they have to obtain their vitamin A primarily from dietary β-carotene, which is converted into retinol within the enterocytes [5]. Independently from provitamin A activity, β-carotene can be absorbed intact in the gut and is considered to have positive effects on reproductive performance and immune response in cattle [6]. All green forages are rich in β-carotene and thus provide a high vitamin A value mainly during pasture conditions [7]. Low concentrations of vitamin A in blood resulting from dietary β-carotene restriction or lacking in vitamin A supplementation and are related to an increased risk of disease [8]. Neonate calves are particularly prone to develop symptoms of vitamin A deficiency, because they have low blood vitamin A concentrations and insufficient hepatic vitamin A stores at birth [9]. Low hepatic vitamin A storage is associated with stillborn and postnatal calf losses as a consequence of an insufficient vitamin A intake of the mother cows especially during non-grazing conditions [10]. In addition, cases of dermatopathy associated with hypovitaminosis A were recently described in juvenile Angus calves, which could be successfully treated with parenteral vitamin A supplementation [11]. Thus, the measurement of blood retinol concentrations in mother cows as well as in neonate and juvenile calves is a prerequisite for evaluation of their health status.

Low vitamin A plasma concentrations are generally found in Japanese Black cattle and are associated with an increased intramuscular or marbling adipose deposition without influencing other adipose depots [12–15]. Thus, to optimize the desired intramuscular fat marbling that increases the carcass value, Japanese Black fattening cattle are provided with a high concentrate diet deficient in vitamin A or β-carotene. It was suggested that optimal marbling scores are obtained when animals are kept very close to health deteriorating low vitamin A plasma concentrations in a range of 100–150 μg/L (40–50 IU/dl) over a period of usually 12 month during the fattening period [16]. However, in Japanese Black cattle low vitamin A plasma concentrations have been reported to be associated with an impaired immune function [17] and increased morbidity and mortality [18]. Therefore, blood vitamin A levels in Japanese Black cattle needs to be carefully monitored and nutritionally controlled. Additional reduced vitamin A plasma concentrations are also observed as a consequence of the acute phase reaction e.g. during inflammation or around parturition [19–21]. Thus, due to the restrict feeding management, the induced vitamin A deficiency in Japanese Black cattle can cause a serious health problem if not monitored and nutritionally controlled properly.

Under practical management conditions, vitamin A deficiency is difficult to diagnose and marginal deficiency may exist without obvious deficiency symptoms. Measurement of retinol in blood plasma is an appropriate marker to assess vitamin A status but it is currently a time consuming and cost-intensive multi-step procedure. Specific equipment is needed to perform extraction and centrifugation or saponification. Vitamin A is finally measured by colorimetric methods [22] or high performance liquid chromatography (HPLC) [23]. All preanalytic steps and the final analyses have to be performed in qualified analytical laboratories by highly trained personal. Alternatively suggested methods such as thin layer chromatography are similar complicated and not very reliable [24]. So far, however, no method has been described which is also able to measure vitamin A directly at cow-side in the barn or even determining vitamin A directly from whole blood prior to centrifugation.

This study presents data for a new, fast, easy to perform and laboratory-independent test to measure retinol in whole blood, plasma or serum. Results are compared with those obtained by HPLC as the reference method at two independent laboratories. The new assay is based on a separation and extraction principal published for the determination of β-carotene in blood plasma [25] as well as human and cow milk [26, 27].

The comparison was performed at two independent laboratory sites at the University of Potsdam, Germany (lab DE) and at the Obihiro University, Japan (lab JP).

The concentrations of retinol in plasma or serum as determined by HPLC and the new method ranged from 0.033 mg/L to 0.532 mg/L for plasma in Lab DE and from 0.043 mg/L to 0.360 mg/L for serum in Lab JP (Additional file 1: Table S1). No significant differences were observed between both methods at two different laboratories for the average values obtained (HPLC vs. new method: 0.320 ± 0.047 mg/L vs. 0.333 ± 0.044. mg/L; mean ± SD in Lab DE and 0.240 ± 0.096 mg/L vs. 0.241 ± 0.069 mg/L; mean ± SD in Lab JP).

The analytical advantage of the new innovative test method is to measure retinol from whole blood directly. This avoids the time consuming and limiting preanalytical step of centrifugation to obtain plasma or serum. In consequence, the new assay can be directly used in the barn. When the subset of 10 whole blood samples were assayed for retinol by the new test method and compared to plasma levels assessed by HPLC in Lab DE, a similar good agreement was found as for the use of plasma (HPLC plasma: 0.353 ± 0.084 mg/L vs. new test plasma: 0.340 ± 0.063 mg/L vs. new test whole blood 0.341 ± 0.064 mg/L; mean ± SD; see Additional file 2: Table S2). There were significantly correlations between concentrations of retinol measured in whole blood by the new method and retinol measured in plasma by HPLC (r² = 0.84; P < 0.001) as well as retinol measured in plasma by the new method (r² = 0.87; P < 0.001). The average packed cell volume of the 10 whole blood samples in this study was 31.7 ± 2.7% (mean ± SD).

The comparison between plasma or serum samples assayed with the two different methods in Lab DE (plasma) and Lab JP (serum) indicated a strong correlation between methods (Fig. 1a, b). Results show that the novel cow-side test for vitamin A correlated very well with HPLC analysis (r² = 0.876 and 0.870, Lab DE and Lab JP, respectively, both P < 0.001). The calculated differences in retinol concentrations obtained by HPLC and fluorometer were not significantly different (P-values >0.05 for both Lab DE and Lab JP) indicating an acceptable level of agreement between the two assay methods. Based on the Bland-Altman plot (Fig. 1c, d) no systematic error and a good level of agreement occurred between the two methods and only 4% of the differences in measured values felt outside the 95% acceptability limits.

Based on multiple determinations of selected samples the intra-assay and inter-assay precision were calculated for each method (Lab DE). In average, the coefficient of variance (CV) was in an order of magnitude generally accepted for average analysis of this kind. The CV was slightly higher in the HPLC method compared to the new method (5.3% vs. 2.3%, respectively).

The comparison of the new cow-side test (iCheck™ FLUORO) to measure vitamin A in cattle whole blood, blood plasma or serum showed a very good agreement, reliability and accuracy with the cumbersome, time-intensive and expensive HPLC analytical technique. The assay allows the determination of vitamin A directly in whole blood within 5 min even at cow-side at the barn without a further sample preparation. Because all chemicals are sealed in the vial, the staff is not directly exposed to potentially hazardous organic chemicals. Additionally, the limited volume of organic solvent necessary is positive with regard to environmentally and ecologically critical waste. The novel test realizes important aspects of a cow-side assay: it is sensitive, specific and rapid. Finally, the test delivers clear end-point results which can help to optimize nutritional interventions in cattle.