Reactive, Not Episodic: When AuDHD Mood Cycling Gets Diagnosed as Bipolar Disorder

A client arrives with a bipolar II diagnosis that is nine years old. She has tried three mood stabilizers. The first flattened her; the second she stopped over weight gain; the third she is still taking, without much conviction. Her chart says partial response in one place and treatment-resistant in another. She has been told, more than once and in more than one tone, that she is not trying hard enough.

She describes her worst weeks. They cluster, she says, around the end of the school term when her kids' schedules fall apart, around office moves, around the month her manager introduced a rotating desk policy. Her best weeks are the ones where nothing changes.

Nobody had asked her that. Nine years of treating the mood, and nobody had asked what the mood was tracking.

She is autistic. She was assessed at thirty-nine, by a colleague, after her daughter's diagnosis raised a question that had been sitting unasked for four decades.

This vignette is a composite. It does not describe any individual client.

It would be convenient to treat this as a story about careless clinicians. It is not. The overlap between AuDHD presentation and bipolar spectrum presentation is genuine, and the diagnostic instruments we use are not built to separate them.

Consider what the two share. Both involve periods of elevated energy and unusual productivity. Both involve periods of collapse, low mood, and reduced function. Both involve sleep disruption. Both involve emotional intensity that observers describe as disproportionate. Both involve impulsivity under load. Both carry elevated rates of depression, anxiety, and substance use. Both are associated with heightened sensitivity to rejection.

Now consider how the assessment usually runs. A screening instrument asks whether the person has experienced a period of unusually elevated mood and energy lasting several days. The AuDHD person, who has spent a lifetime in cycles of hyperfocus followed by depletion, says yes — because they have. The instrument asks whether they have experienced periods of low mood, fatigue, and withdrawal. They say yes — because autistic burnout is exactly that. Two yeses, and the pattern is complete.

What the instrument does not ask is what surrounded those periods. And the surrounding context is where the entire differential lives.

The temptation is to distinguish these by intensity — to imagine that bipolar mood states are bigger and AuDHD variability is smaller. This is wrong, and clinically it is the error that does the most damage, because AuDHD mood states can be enormous. A person in the third day of a shutdown after an unannounced office reorganization is not experiencing a mild version of anything.

The difference is structural.

Duration and course. Bipolar mood episodes have a defined shape. A manic episode runs at least a week; a hypomanic episode at least four consecutive days. The episode has a beginning, a middle, and an end, and it does not particularly care what is happening in the person's life while it runs. AuDHD mood variability is measured in hours to days, and its shape is dictated by the load that produced it.

Relationship to baseline. In bipolar disorder, the episode is a departure. There is a baseline to return to, and between episodes people often do return to it. In AuDHD, the variability is the baseline. It is trait-like and lifelong — consistent variability rather than episodic variability. When you take a careful developmental history, you are not looking for the first episode. You are looking for whether there was ever a period without the pattern.

Contingency. This is the crux. Bipolar episodes are endogenous — they arise from internal processes and can appear with no external trigger at all. AuDHD mood states are contingent on identifiable load: sensory environment, transition density, demand pressure, executive burden, social recovery debt, interoceptive state, sleep, illness, hormonal cycle. There is nearly always an antecedent. It is just that nobody asked about it, and the person did not volunteer it, because to them the antecedent is unremarkable — it is simply what Tuesday is like.

Sleep. The most discriminating single symptom, and the one most often collected badly. In mania and hypomania, the cardinal feature is decreased need for sleep: the person sleeps three hours and feels genuinely rested and energized. In AuDHD, sleep is frequently disrupted — delayed onset, revenge bedtime procrastination, sensory interference, circadian irregularity — but the person is tired. They are running on insufficient sleep and feeling it. "I was up until four and I was fine" points one direction. "I was up until four and I was wrecked but I couldn't stop" points the other.

The quality of the high. Hypomania typically involves expansiveness — inflated self-regard, grandiosity, an unusual sense of one's own capability, goal-directed activity across multiple new domains. AuDHD hyperfocus is typically narrow, absorbing, and specific. It fixes on one thing. The person is not expansive; they are captured. They often cannot describe the period as pleasurable so much as involuntary, and it is frequently followed by a crash proportional to its length.

Three mechanisms keep this misdiagnosis stable once it forms.

Masking obscures the developmental history. The adult who was missed as a child was usually missed because they were compensating well. When a clinician takes a history and asks whether there were early signs of autism, the person says no — because the signs were internal, and because they have spent thirty years being told that the effort they were expending was ordinary and everyone felt that way. The developmental history that would establish trait-like variability comes back empty, and the episodic reading survives.

Partial medication response is ambiguous. Mood stabilizers are not inert in an AuDHD nervous system. They blunt affect, and blunted affect reduces the amplitude of reactive mood states. A person may report that things are somewhat better on lithium or lamotrigine — and this is read as diagnostic confirmation. It is not. Reduced amplitude is not the same as resolved mechanism. The person still cycles; they just cycle less loudly, at the cost of accessing their own emotional range.

The label becomes the explanation. Once bipolar is in the chart, subsequent variability is interpreted through it. A shutdown after a schedule change is recorded as a depressive swing. A hyperfocus week is recorded as hypomania. The antecedents stop being collected, because the diagnosis already accounts for the phenomenon. The chart becomes internally consistent and externally wrong.

And the distribution is not random. The people who reach adulthood undiagnosed are disproportionately those whose presentation did not match the pattern clinicians were trained to look for — women and AFAB people, transgender and gender-expansive people, people of color, and anyone whose distress was read as behavioral rather than neurological. The misdiagnosis pipeline runs through the same channels every other diagnostic disparity runs through.

The reframe is small and the yield is large. Stop asking primarily about the mood. Start asking about what surrounded it.

• What happened in the seventy-two hours before this started? Not "was there a stressor" — that invites a no. Ask concretely: what changed in the schedule, the environment, the noise level, the lighting, the number of people, the number of transitions, the amount of unstructured social contact.
• When you were at your best, what were the conditions? People will describe stability, predictability, and low sensory load without knowing that is what they are describing.
• During the high periods — did you need less sleep, or did you get less sleep? Ask it as an either/or. The answer is usually immediate and unambiguous.
• Was the high about one thing or about everything? Narrow and absorbing points toward hyperfocus; broad and expansive points toward hypomania.
• Has there ever been a stretch of years without this pattern? A true baseline period supports an episodic reading. Its absence supports a trait-like one.
• Who first raised the mood question, and what were they seeing? Frequently it was a partner or parent reporting on observable dysregulation, not the person reporting internal mood states.

None of these are exotic. They take perhaps ten minutes. They are simply not in most intake protocols, which are built to characterize symptoms rather than to locate them in an environment.

An argument this clean deserves its own cross-examination.

Comorbidity is real and common. Autistic people develop bipolar disorder. ADHD people develop bipolar disorder, at rates well above the general population. The existence of a plausible alternative explanation does not exclude the original diagnosis, and a clinician who treats this differential as a one-way door will do harm in the other direction — dismissing genuine bipolar illness in a neurodivergent person as "just sensory," which is its own well-documented failure.

The literature is contaminated in both directions. If AuDHD adults have been systematically absorbed into bipolar cohorts for decades, then some of what we know about "bipolar" populations describes them. That cuts against confident claims from that literature — including some I have made here. Treat the mechanism as better established than the prevalence estimates.

Reactive does not mean minor, and endogenous does not mean severe. I have drawn this distinction on architecture precisely because severity does not track it. A reactive mood state can be disabling. An endogenous one can be mild.

Bipolar disorder carries mortality risk that this differential must not obscure. Untreated bipolar illness is dangerous. If the differential is wrong in the direction of under-diagnosis, the consequences are more serious than if it is wrong in the other direction.

Which leads to the thing that matters most:

If the reactive reading is correct, the treatment target moves — from stabilizing an internal cycle to reducing the load that drives an external one.

That means sensory accommodation as clinical intervention rather than lifestyle advice. Transition scaffolding and advance notice as a negotiated accommodation at work and at home. Recovery time budgeted after social demand rather than treated as laziness. Executive supports. Interoceptive work, because many AuDHD adults cannot detect their own load until it has already tipped. Where ADHD is present and appropriate, stimulant treatment — which is frequently withheld for years on the strength of the bipolar label.

And it means something less clinical and harder to bill for. The person who has spent nine years being described as treatment-resistant has usually concluded something about themselves. That they are difficult. That they failed at getting better. Rewriting that conclusion — locating the failure in the formulation rather than in the person — is often the intervention that does the most work.

Mood stabilizers do not reduce sensory load. They were never going to. That is not a failure of the person taking them.

Content on this site is for educational and informational purposes only and does not constitute medical advice, diagnosis, or treatment. Nothing here creates a therapist–client relationship. If you are in crisis, call or text 988 (Suicide & Crisis Lifeline) or go to your nearest emergency room.