including perhaps overall withdrawal severity. However, including the maximum CIWA score for each individual obtained within the first couple of days of acute withdrawal as a potential mediator did not change the results (data not shown). Furthermore, analyzing anxiety and depression ratings measured subsequent to acute withdrawal (on day 9, resulting in a decreased sample size of 234 subjects due to missing data), did not alter the results not contribute significantly to the outcome (data not shown).
