The problem with DSM dependence

This NIAAA feature points out a big problem with the DSM diagnosis alcohol dependence. The diagnosis is intended to capture people with chronic problems and the NIAAA article suggests that as many as 70% have one episode that lasts less than 4 years and that 75%  of those who "recover" do so without any treatment. This leaves 30% having more than two or more epsiodes. The former group appears to be people who go through a period of heaving drinking (most often as young adults) and then moderate or stop drinking on their own.

This begs all sorts of questions, such as:

• Should these two groups be in one diagnostic category? 
• I what ways are these categorically different types of drinkers? Do they respond to different treatments? Do they require different treatment goals? (abstinence vs. moderation) 
• Do these people who "recover" think of themselves of recovering? Should they?
• Are there ways for clinicians to distinguish between these two types? Are the people with more severe forms more likely have the chronic form? Are there patterns in terms of the diagnostic criteria that are met for the 2 groups? (I suspect that the loss of control criteria are more frequently met for the people with the chronic type.)
• When we read stories about the number of people who need treatment and don't receive it, are those 70% included? (yes) What does this mean for those kinds of statistics?

[via Adi Jaffe]