The 7 UI patterns that quietly exclude ADHD users

A row of hero slides that automatically advances every 4–8 seconds. Sometimes with pause-on-hover, sometimes without. Almost always with the next slide already animating in before the previous one's headline has been read.

Sustained reading depends on stable visual context. Auto-rotation introduces involuntary motion in the user's foveal field, which the brain's salience network registers as a potential threat or novelty signal^[3]. For neurotypical readers this is a minor annoyance. For ADHD users, who already have reduced top-down control over attentional capture, the carousel resets the reading process every time it advances.

Nielsen Norman Group's carousel research shows neurotypical users almost universally interact only with the first slide, and that's without attentional differences in play^[5]. The pattern's stated benefit (showing more content in less space) is mostly fictional.

Critical navigation or information hidden behind a hover. The menu only opens on :hover; the tooltip only appears when the cursor lingers. On mobile, where hover doesn't exist, behaviour degrades unpredictably.

ADHD frequently co-occurs with motor planning differences and impaired response inhibition^[6]. The user does not patiently explore a UI by hovering deliberately over each region, they click on what looks clickable. Hover affordances are invisible until investigated, and ADHD users are statistically less likely to invest exploration time before acting.

This is also a WCAG 2.2 concern. Success Criterion 2.5.5 (Target Size) and 1.4.13 (Content on Hover) explicitly require alternatives to hover-only revelation^[4].

Confetti when a form submits successfully. A bouncing checkmark on "Saved." Sliding panels and exaggerated transitions used to communicate state change rather than colour or text.

The salience network is the brain region that decides which environmental signals deserve attention. ADHD is associated with over-active salience network responses to motion^[3]. Decorative animation does what your interface intended (signals success) and also what it didn't (drains attention from the next step).

This compounds with vestibular sensitivity, a meaningful subset of ADHD users experience light-headedness or nausea from heavy parallax and zoom transitions^[4]. WCAG 2.3.3 (Animation from Interactions) explicitly addresses this.

Pulsing grey rectangles that approximate the shape of incoming content. Often persist for 1–5 seconds while data fetches. No indication of how long the wait will be or whether progress is being made.

The Doherty Threshold, the response time below which user productivity climbs sharply, is 400ms^[7]. Above that, users experience friction. ADHD users experience that friction more acutely because time perception is distorted: short waits feel long, and the absence of progress information amplifies that distortion^[3].

A skeleton loader that pulses for three seconds without telling the user how many of ten items have loaded creates uncertainty. Uncertainty in turn drains working memory because the user must hold the question "is this still loading?" alongside whatever they came here to do^[2].

Content loads endlessly as the user scrolls. No pagination, no "you've reached the end," no sense of position in the list. The footer becomes unreachable.

ADHD time-perception deficits mean users genuinely cannot tell whether they've been scrolling for two minutes or fifteen^[3]. Without spatial anchors (a scrollbar that means something, a "page 4 of 12" indicator, a visible end), the experience becomes a slot machine, the dopamine reward is real, the cost is invisible, and the user often emerges having lost time they did not intend to spend.

Working memory makes this worse: ADHD users have a harder time keeping a plan active (e.g., "I came here to find one specific item")^[2] while their attention is captured by an unending stream of novel content. The plan dissolves; the scrolling continues.

The hamburger menu on a desktop layout with screen real estate to spare. Accordion-heavy account settings where every label is collapsed by default. Action buttons that only appear on hover-over-row in a table.

Working memory in ADHD has measurably smaller capacity, the Martinussen meta-analysis found large effect sizes for both verbal and spatial working-memory deficits^[2]. Every "click to reveal" pattern adds an item the user must hold mentally: "the setting I need is behind that accordion."

Hick's Law tells us decision time scales logarithmically with options^[8]. Hiding options doesn't reduce decision time, it just defers the cost. The decision still has to happen; it now happens under conditions of additional uncertainty (am I in the right place?) and additional working-memory load (where was that thing?).

A wizard with no "Step 3 of 7" indicator. The user can move forward and back, but has no idea how many steps remain. No option to save and resume. Validation errors send them back to a state they no longer remember.

This is the convergence of every prior pattern. Working memory burden (what did I just put in?)^[2]. Time-perception distortion (how long is this taking?)^[3]. Inhibitory-control demand (resist the urge to close the tab)^[6]. Sustained-attention demand (don't get distracted before submit)^[9].

Abandonment rates for un-indicated multi-step forms are 2–3× higher than for forms with progress markers, across the general population^[10]. The effect is significantly larger for ADHD users, though precise figures from ADHD-specific UX studies are still scarce, an open invitation for further research.