Subclinical ketosis is defined as high serum ketone body concentrations without observed clinical signs. Subclinically affected cows are at increased risk of clinical ketosis and displaced abomasum and are also less fertile than animals with normal serum ketone body concentrations. Furthermore, they appear to have reduced milk production. Determination of serum BHB concentrations is considered the best means of detecting and monitoring subclinical ketosis, because the cowside tests mentioned above are insufficiently sensitive and specific in detecting subclinical increases in serum BHB concentrations. Serum concentrations may be determined spectrophotometrically by traditional clinical laboratory means. The BHB concentrations in blood or serum samples are reasonably stable, thus rigorous sample handling precautions are not necessary for transporting the specimens to the laboratory. The test is sensitive to hemolysis, however, so hemolysis should be avoided during sample collection and serum should be separated from the clot prior to shipment to the laboratory.

In addition to laboratory determination by spectrophotometry, handheld devices manufactured for monitoring blood ketone body concentrations in diabetic human patients have been evaluated for use in monitoring subclinical ketosis in cows. These instruments use whole blood rather than serum for BHB determination, making them particularly practical for on-farm use. The whole blood BHB concentration is very close to the serum concentration, so the interpretation of results obtained from either the handheld device or laboratory analysis is similar.

Diagnosis of subclinical ketosis requires definition of a concentration above which cows are considered to be subclinically ketotic. Concentrations between 1,000 µM (10.4 mg/dL) and 1,400 µM (14.6 mg/dL) are used, with 1,400 µM appearing to be used most commonly. Recommended strategies for herd-level testing are to test at least 12 animals in the first 60 days of lactation. If >10% are subclinically ketotic, it should be considered evidence of a herd-level problem and prompt a review of nutritional management.

Last full review/revision July 2011 by Thomas H. Herdt, DVM, MS, DACVN, DACVIM