Approach bias modification
Approach bias modification has been studied in 2 papers since 2024: 2 trials in people; 1 in the past year. The newest paper, Mar 28, 2026: “The effects of approach bias modification on smoking cue-reactivity in individuals who smoke: A randomized controlled fMRI study”.
2 papers study it7 findingsUpdated
“with past-week standard drinks significantly decreasing by 18% (p < 0.001)”
None of its findings points the other way.
- 2 independent labs in 2 countries
- Funders: Deutsche Forschungsgemeinschaft and 8 more
- Human trial2
Studies in people
2 studies gave approach bias modification to people and measured what changed.
no change: Has no effect on smoking abstinence26.2% vs 16.7% vs 23.1%
no change: Has no effect on smoking cue-reactivity
Randomized trial117 peopleBoth18–70 yearsseven daily sessions
Individuals who smoke with moderate-to-heavy dependence
Scientific reports, Mar 28, 2026
down: Decreases alcohol consumption-18%, p < 0.001
up: Increases adherence to alcohol guidelinesfrom 25% to 41% and 6% to 28%, p < 0.001
down: Decreases alcohol dependencep < 0.001
Trial289 people≥55 years4 weeks
Midlife adults; Older adults
Drug and alcohol review, Mar 5, 2024
What did studies of Approach bias modification find?
5Studies of Approach bias modification measured 5 outcomes; the most studied are smoking cue-reactivity (1 study), smoking abstinence (1) and alcohol dependence (1).
Smoking cue-reactivity
no changein humans1
1 study
Smoking cue-reactivity
no changein humans1
Approach bias modification has no effect on smoking cue-reactivity in humans.
“Contrary to our hypothesis, omnibus tests revealed no significant group×time interactions on smoking cue-reactivity in any ROI (ps ≥ 0.235; see Table and supplementary results Fig. for boxplot visualization), indicating comparable changes in cue-reactivity across groups.”
Human trialseven daily sessionsn = 117
Scientific reports
Approach bias modification has no effect on smoking cue-reactivity in humans.
“Results showed no significant group×time interactions on cue-reactivity”
Human trialtobacco dependencecomputerized trainingreward-related brain regionsseven sessions45.3% femalesn = 117
Scientific reports
Smoking abstinence
no changein humans1
1 study
Smoking abstinence
no changein humans1
Approach bias modification has no effect on smoking abstinence in humans (26.2% vs 16.7% vs 23.1%).
“Regarding group differences in abstinence rates, no significant effects were found at t1 (short-term abstinence; whole sample: 56 of 117 [47.9%], TAU+ApBM: 19 of 42 [45.2%], TAU+Sham: 19 of 36 [52.8%], TAU-only: 18 of 39 [46.2%]; χ2[2] = 0.51, p =.775) and at t2 (long-term abstinence; whole sample: 26 of 117 [22.2%], TAU+ApBM: 11 of 42 [26.2%], TAU+Sham: 6 of 36 [16.7%], TAU-only: 9 of 39 [23.1%]; χ2[2] = 1.04, p =.594”
Human trialseven daily sessionsn = 117
Scientific reports
Approach bias modification has no effect on smoking abstinence in humans.
“ApBM did not enhance abstinence rates.”
Human trialtobacco dependencecomputerized trainingseven sessions45.3% femalesn = 117
Scientific reports
Alcohol dependence
downin humans1
1 study
Alcohol dependence
downin humans1
Approach bias modification decreases alcohol dependence in humans (p < 0.001).
“significant reductions in drinking days, mean craving and dependence scores (p < 0.001).”
Human trialsmartphone-delivered and personalized4 weeksn = 289
Drug and alcohol review
Adherence to alcohol guidelines
upin humans1
1 study
Adherence to alcohol guidelines
upin humans1
Approach bias modification increases adherence to alcohol guidelines in humans (from 25% to 41% and 6% to 28%, p < 0.001).
“significant increase in the proportion of drinking within recommended single-session and weekly guidelines post-training (from 25% to 41% and 6% to 28%, respectively, p < 0.001)”
Human trialsmartphone-delivered and personalized4 weeksn = 289
Drug and alcohol review
Alcohol consumption
downin humans1
1 study
Alcohol consumption
downin humans1
Approach bias modification decreases alcohol consumption in humans (-18%, p < 0.001).
“with past-week standard drinks significantly decreasing by 18% (p < 0.001)”
Human trialsmartphone-delivered and personalized4 weeksn = 289
Drug and alcohol review