Additionally, we do not have data on formal treatment, AA attendance, and alcohol abstinence during years 2, 4, and 6 of this 7-year longitudinal study. It is possible that events and behaviors in the years for which we do not have data were associated with behavior years 3, 5 or 7, potentially resulting in an overestimation of the associations.
