Association between smoking-related attentional bias and craving measured in the clinic and in the natural environment

Objective
Previous studies have investigated the association between attentional bias and craving in laboratories, but ecological momentary assessment (EMA) may provide ecologically valid data. This study examines whether clinic-measured attentional bias is associated with noticing smoking cues, attention to smoking, and craving assessed by EMA and whether EMA-assessed cues and attention to smoking are associated with craving.
Method
Secondary analyses of clinical trial data involving 100 cigarette smokers attempting to quit assisted by behavioural support and nicotine patch treatment. Two weeks prior to quitting, participants completed attentional bias assessments on visual probe and Stroop tasks. Participants carried personal digital assistants for seven weeks thereafter, which administered random assessments of smoking cues, attention towards smoking, and craving.
Results
Participants completed 9271 random prompt assessments, averaging 3.3 prompts/day. There was no evidence that clinic-measured attentional bias was associated with cues seen (visual probe: OR=1.00, 95%CI=0.99, 1.01; Stroop: OR=1.00, 95%CI=0.99, 1.00), attention towards smoking (visual probe: OR=1.00, 95%CI=0.99, 1.02; Stroop: OR=1.00, 95%CI=0.99, 1.00),
or craving (visual probe: OR=1.00, 95%CI=0.99, 1.02; Stroop: OR=1.00, 95% CI=0.99, 1.01). EMA responses to seeing a smoking cue (OR=1.94, 95%CI=1.74, 2.16) and attention towards smoking (OR=3.69, 95%CI=3.42, 3.98) were associated with craving. Internal reliability was higher for the Stroop (α=0.75) than visual probe task (α=0.20).
Conclusions
In smokers attempting cessation, clinic measures of attentional bias do not predict noticing smoking cues, focus on smoking, or craving. However, there are associations between noticing smoking cues, attention towards smoking, and craving when assessed in the moment, suggesting that attentional bias may not be a stable trait.