This invention relates to a method for detecting the presence of or likelihood of a patient of developing occult pancreatic beta cell dysfunction, and a method for detecting the presence of or likelihood of a patient of developing clinically significant post-prandial hyperglycemia. The methods involve (a) measuring a level of alpha-hydroxybutyrate (AHB) in a single fasting baseline biological sample of the patient; (b) comparing the level of AHB in the single fasting baseline biological sample to a reference AHB level; and (c) determining the presence of or likelihood of developing the disorder in the patient based on the comparison in step (b). An increased AHB level at fasting baseline indicates that a normoglycemic, normo-insulinemic and/or non-dyslipidemic patient has developed or has an increased likelihood of developing occult pancreatic beta cell dysfunction. An increased AHB level at fasting baseline and an elevated glucose level of at least about 155 mg/dL at 30 minutes and/or 1 hour indicates that a normoglycemic, normo-insulinemic and/or non-dyslipidemic patient has developed or has an increased likelihood of developing clinically significant post-prandial hyperglycemia.