Yes, absolutely.
> For example, the current guideline is to diagnose a patient as diabetic if their uncontrolled hemoglobin A1c is over 6.5%.
Several other objective tests can be used to diagnose diabetes. The disease could present with characteristic symptoms which match diabetes. The presence of comorbidities such as obesity also help with diagnosis.
> But the reality is that it's a spectrum with no clear dividing lines, significant fluctuations from day to day, individual genetic variations, etc.
This spectrum is called pre-diabetes. At some point, the risk becomes unacceptably high for the patient and treatment is indicated. This point has been determined through research.
Yes, blood glucose does vary during the day depending on the person's metabolic state. This variability is well understood. Normal and abnormal is well-defined. Fasting blood glucose > 126 mg/dl is evidence of diabetes mellitus. Blood glucose 2 hours after ingesting 75 g of glucose > 200 mg/dl is evidence of diabetes mellitus.
Likewise depression is just a convenient label for a bundle of loosely correlated symptoms. Some patients are clearly having a rough time and need treatment. Because it's impractical to deal with subtle differences between patients at scale, psychiatrists have defined some subjective criteria to decide who should be treated as depressed.
https://www.medscape.com/answers/286759-14692/what-are-the-d...