The issue is that there is a measurement problem with many depression symptoms. You can have multiple people independently measure the blood glucose level of a

The issue is that there is a measurement problem with many depression symptoms. You can have multiple people independently measure the blood glucose level of a patient and get the same result. And this result compares with other patients. With depression, you almost always have to rely on self-reporting or subjective evaluation by a health professional. One person saying they are depression might mean a whole lot more than another person saying it. It's not that you can't make measurements that are useful, its just that the error bars are so much larger.