By analyzing the results from four completed NIDA-funded studies, the major goal of this revised R03 proposal
is to assess whether long-term smoking relapse is predicted by decreases in positive affect (PA) and
cognitive functioning (CF) and by slow recovery of PA and CF during the first days and month of smoking
abstinence in those who maintain abstinence for at least 31 days. A second study goal is to replicate findings
(Zuo et al., 2017) that slow dissipation of negative affect (NA) during the first days of abstinence predicts both
short- and long-term relapse to smoking. The third goal of the study is to use data from the four studies to
generate a clinically useful risk-for-relapse scale (Short Cognitive-Affective Relapse Risk Scale [SCARRS])
based on the most predictive withdrawal-symptom items from the studies. While NA was assessed as a
predictor of relapse in the Zuo et al. study, PA and CF were collected but not assessed as potential predictors
in any of the four studies, each of which assessed smoking withdrawal symptom (SWS) trajectories for a
minimum of 31 days of incentivized, biochemically verified smoking abstinence, and collected relapse
information at 3, 6, 9, and 12 months. An advantage of assessing these relationships in studies that provided
incentives for maintaining abstinence during the first 31+ days is that individuals who normally would have
relapsed and therefore not have been included in the assessed sample are maintained in the abstinent state,
something that avoids the problem of those with the most severe SWSs relapsing and not being included in the
study population. Exclusion of such individuals typically results in underestimates of SWS severity and duration
and minimizes the ability to assess the relationship of SWS to relapse in those who maintain initial abstinence.
In addition, a sizable portion of smokers relapse after 31+ days. If slow dissipation of NA and/or slow recovery
of PA and/or CF during early abstinence are found to predict a persistent risk for later relapse, these variables
may be valid targets for interventions to increase long-term cessation in individuals with refractory SWSs. To
facilitate clinical adaptation, we will generate a brief SWS predictor of risks for relapse based on the statistical
identification of items that most effectively predict relapse from the standard lengthier measures of SWSs used
in the four studies. To probe the earliest time window for individualized intervention, we will assess whether the
rate of dissipation of SWSs across days 3 through 31 of abstinence can be established as markers of
significance for relapse; and specifically, whether greater increases in SWSs from baseline to the third day of
abstinence, slower resolution of these SWSs from day 3 to day 6 and to day 12 and 31 of abstinence are each
predictive of relapse, even after controlling for craving. Growth curve and logistic regression analyses for data
combined across the studies will be used to assess associations of an individual’s SWS trajectory parameters
and craving with the abstinence outcomes at 3-month and 1-year postquit. A finding that the very brief (e.g.,
12-20 item) SCARRS can identify individuals at high risk for relapse could lead to its widespread clinical use.