ADHD-associated risk taking is linked to exaggerated views of the benefits of positive outcomes

The assumption is this: people with attention deficit hyperactivity disorder take more risks because they do not feel danger the way others do. They underestimate the downside. They are disinhibited. The fear signal is turned down. That assumption has shaped decades of research and intervention. Shoham and colleagues looked directly at it — and found it is wrong. The fear signal, it turns out, is completely intact. It is the reward signal that is different. That is the finding. Now let's walk through how they got there, and why the distinction matters. Shoham and colleagues approached ADHD-related risk taking through behavioral economics, specifically through what is called a psychological risk-return model. The core idea is that any risky decision involves two separate subjective components: how dangerous an option looks — perceived risk — and how attractive its upside looks — perceived benefit. These are not the same thing, and they are not simply mirror images of each other. One is about estimating harm; the other is about craving reward. The model also separates raw perceptions from what the researchers call attitudes — how strongly perceived risk repels you, and how strongly perceived benefit draws you in. Stated as an equation you can hear: preference for an option equals a benefit-weight times perceived benefit plus a risk-weight times perceived risk. The question for ADHD research is which of those terms is doing the work. To test this, the team recruited two hundred forty-four adults — one hundred thirty-two women and one hundred twelve men, with a mean age just over thirty-four — and had them complete two instruments in Hebrew. The first was the Adult ADHD Self-Report Scale, or ASRS, an eighteen-item questionnaire with strong reliability. The second was the Domain-Specific Risk-Taking scale, known as DOSPERT. This is where the design gets elegant. DOSPERT does not just ask whether you would do something risky. It asks three separate questions for each of thirty behaviors: how likely are you to do this? How much benefit do you see in it? How much risk do you see in it? Those questions are rated on seven-point scales, and they cover five behavioral domains — health and safety, financial, ethical, recreational, and social. One example item is driving without wearing a seat belt. By splitting likelihood, benefit perception, and risk perception into separate ratings, the researchers could see exactly which dimension tracked with ADHD symptoms. They also computed individual attitude scores from each person's pattern of responses, capturing how much their choices were driven by benefit-seeking versus risk-avoidance. The results were striking in their asymmetry. ADHD symptom scores on the ASRS correlated positively with self-reported risk taking — a Spearman correlation coefficient of 0.27 — and correlated positively with perceived benefits from risky behaviors, at 0.29. The correlation between ADHD symptoms and risk perception was essentially zero: correlation of negative 0.005, which was not statistically significant. That finding held for both symptom dimensions separately. Inattention scores correlated with risk taking and benefit perception but not with risk perception. Hyperactivity and impulsivity scores showed the identical pattern. Both dimensions exhibited the same structure. The danger signal is intact. The reward signal is amplified. To understand what that means in practice: someone high in ADHD symptoms is not looking at a risky behavior and thinking it is safe. They see the same danger everyone else sees. They just think the upside is much, much better. The mediation analysis is where this gets formally tested. Mediation asks: does the path from ADHD symptoms to risky behavior run through a third variable? The team used Hayes' PROCESS macro, Model 6, with bootstrapped confidence intervals and age and gender as covariates. The full model accounted for fifty-seven point three percent of the variance in risk behavior. The standardized indirect effect of ADHD symptoms on risk taking mediated by perceived benefit came in at 0.18, with a ninety-five percent confidence interval running from 0.12 to 0.22 — clearly significant. The indirect effect through perceived benefit attitude was also significant but smaller and negative, at negative 0.04. By contrast, the indirect effect through risk perception was 0.004, with a confidence interval that crossed zero. The indirect effect through risk attitude was similarly null. The data are consistent: the road from ADHD symptoms to risky behavior runs through how good the reward looks, not through any reduction in perceived danger. There was also a residual direct effect of ADHD symptoms on risk taking — a standardized coefficient of 0.11 — meaning benefit perception explains much of the relationship but not all of it. Something else may also be operating, though the paper does not pin down what. The domain-by-domain breakdown is worth a moment. In four of the five DOSPERT domains — health and safety, recreational, ethical, and financial — the pattern held cleanly: ADHD symptoms correlated with more risk taking and higher benefit perception but not with risk perception, and benefit perception mediated the link in each. The health and safety domain model explained forty-eight point six percent of variance on its own. The social domain broke the pattern. There, ADHD symptoms correlated with higher risk perception, not higher benefit perception, and risk perception — not benefit — served as the mediator. The authors note this but do not fully explain it; social risks may simply have a different cognitive structure than physical or financial ones. The finding also held when the researchers split the ASRS into its inattention and hyperactivity-impulsivity subscales and ran the mediation separately for each. For inattention, the benefit-perception indirect effect was 0.18; for hyperactivity-impulsivity, it was 0.15. Both subscales told the same story. Now for what this means — and for where the standard thinking goes wrong. If the dominant model were correct — if ADHD risk taking were driven by underestimating danger — then the most logical intervention would be to help people with ADHD appreciate risk more accurately. Warn them more clearly. Show them the consequences. But if the actual mechanism is inflated benefit perception, those approaches are targeting the wrong variable entirely. The person already sees the danger. What they are seeing differently is how good the reward looks. The lever is on the benefit side of the equation. Shoham and colleagues make this point explicitly. Reducing risky behavior in people with high ADHD symptoms requires attention to the perceived attractiveness of risky options, not just to risk communication. That could mean cognitive approaches that examine and reality-test benefit estimates, or external structures that reduce the salience of immediate gains. The study has real limitations worth naming. The sample skewed educated and was not clinically recruited — only about nine percent of participants had a formal ADHD diagnosis; the rest were assessed dimensionally by symptom scores. Risk taking was measured through self-report and hypothetical ratings on the DOSPERT, not through observed behavior or collateral reports. It is possible that people high in ADHD symptoms also over-report the benefits of behaviors they engage in, making the perceived benefit score partly a post-hoc rationalization rather than a prior cause. Mediation analyses cannot fully resolve that directionality question from cross-sectional data. Still, the conceptual contribution is solid. By importing the behavioral economics framework — separating perceived risk from perceived benefit, and separating perceptions from attitudes — Shoham and colleagues were able to ask a more precise question than the field had been asking. And the answer is consistent across multiple analytic approaches, multiple domains, and both symptom dimensions of ADHD. People with higher ADHD symptoms are not taking more risks because the world looks less dangerous to them. They are taking more risks because the rewards look more dazzling. That is a specific cognitive distortion — an inflation of the perceived upside of risky choices — and naming it precisely is the prerequisite for addressing it. Fear-based warnings may be the wrong tool for a bias that lives on the reward side of the ledger entirely. This lecture was created by ennepō. Go to https://ennepo.ai to Discover, Create and Follow the latest research in your field. Read when you can. Listen when you want to.