When you can, scale up: Large-scale study shows no effect of tDCS in an ambiguous risk-taking task

- We aimed to assess whether neuromodulation affects risk taking.
- tDCS applied to the dorsolateral prefrontal cortex does not affect risk taking.
- Previous significant effects are likely to reflect random variation.

BackgroundA wide range of neuroimaging and neuromodulation studies have shown that the dorsolateral prefrontal cortex (DLPFC) plays a pivotal role in decision-making. Of particular interest is the question of its role in decision-making when conditions are uncertain and whether manipulating this neural substrate through neuromodulation changes subsequent risk-taking behaviour. Previous work using the Balloon Analogue Risk Task (BART) suggests that bilateral tDCS stimulation of the DLPFC reduces risk-taking behaviour but unilateral stimulation has no effect. However, participant numbers have been limited and may have biased the estimate of the size of the effect of the stimulation on task performance.Objectives/hypothesisWe aimed to test the robustness and generalizability of these previous findings by using a very similar methodology but with a much larger sample.MethodsDuring both 20- and 30-min tDCS stimulation at 2 mA, we administered the BART to about 200 participants assigned to bilateral DLPFC stimulation of either right anodal/left cathodal, left anodal/right cathodal or sham (Study 1 and Study 2); and to unilateral stimulation conditions (Study 2): right anodal, left anodal or sham with the referent electrode over the contralateral supraorbital region.ResultsIn the first bilateral study, we found that risk-taking was greater for participants in the right anodal/left cathodal stimulation group compared to those who received left anodal/right cathodal stimulation, but not compared to sham. The results obtained in the bilateral and unilateral stimulation protocols implemented in Study 2 yielded no evidence of any effect of stimulation. Combining the data from both studies, we found no statistically significant differences between mean performances of the nine stimulation groups. Indeed, all 95% confidence intervals for the nine means overlapped, suggesting that these randomly vary around a common population mean.ConclusionsThis study showed that there was no detectable effect of tDCS stimulation on risky decision-making under ambiguity, compared to sham stimulation. Hence, using a much larger sample, we did not replicate previous work reporting a reduction in risky decision-making by bilateral stimulation of the DLPFC compared to sham. When the results of our bilateral and unilateral stimulation studies were combined, it emerged that the most likely explanation for the apparent significant results in our bilateral stimulation study was random variation in performance. This outcome is a further reminder of the need for appropriately sized samples to potentially achieve reliable outcomes in brain modulation studies.