self-medication

Self-medication happens when someone uses substances, consciously or unconsciously, to help with something they might not even know is there. In the clinical literature, the term goes back to Edward Khantzian’s 1997 self-medication hypothesis, which proposed that substance use often reflects an attempt to relieve psychological distress rather than random pleasure-seeking¹. The idea has since been extended to neurodivergent adults, where the substances a person reaches for often act on the same systems that work differently in their brain.

Unmet needs and self-medication

Self-medication doesn’t need conscious intent. You don’t have to think of your morning coffee as such for it to be doing the work of self-medication. You could have been drinking five cups a day since university, and not ever wonder why you feel dysfunctional without it. (Spoilers: Because it could be topping up an underlying dopamine system that runs differently for you.)

Many late-identified neurodivergent adults recognise their self-medicating tendencies only in looking back at their lives. For example, when they start collecting evidence for an ADHD assessment and see that copious amounts of coffee, tea, energy drinks or wine & cigarettes had been doing regulatory work all along. Or when they start titrating their stimulant medication and notice that suddenly they ‘need’ fewer cups of tea or coffee to function.

Self-medication is not a neutral term, and has clinical baggage. In psychiatry, self-medication has sometimes been treated as a “warning sign’ or a step toward addiction, which sets up a moral hierarchy where prescribed medication is “treatment” and everything else is a “failure of self-management”. But neurochemistry doesn’t care about morality. Prescribed medication and self-medication act at different points in the same systems.

But whether the substance came from a pharmacy or a coffee shop, it doesn’t change what’s happening at the receptor level.

So if you learn more about it, you can have a new framework for it, and it can open the door to more sustainable options if the current pattern has stopped working, or was never quite working in the first place.

How self-medication works

Regulating anything in our bodies happens through chemistry. Attention, mood, arousal, and social engagement all depend on neurotransmitter activity in specific circuits, which are the same circuits that behave differently in neurodivergent brains.

Dopamine tone can be lower in the prefrontal cortex of an ADHD adult. Sensory input can arrive with less filtering in an autistic nervous system. It is perceptible from the inside as brain fog, restlessness, or overload, and reaching for something that has the chance of shifting the underlying chemistry is a reasonable response to that internal state, so no judgment here.

Prescribed medications and the things people reach for on their own act on the same neurochemistry.

A dopamine neuron produces its output through a chain: amino acid precursors like tyrosine get converted into dopamine by an enzyme (tyrosine hydroxylase) that requires iron to work, and the finished dopamine gets released, received, and cleared away in a controlled cycle. Anything that shifts one link in that chain will shift what actually reaches the receptors. An L-tyrosine supplement adds raw material at the start. Iron enables the enzyme. Methylphenidate slows reuptake, so dopamine stays in the synapse longer³. Caffeine works one step removed, blocking adenosine receptors and indirectly raising dopamine availability⁴. Nicotine acts closer to the release step, prompting release in the same circuits methylphenidate targets⁸. Lisdexamfetamine increases release and inhibits breakdown at the same time. (For more on the dopamine system, see the separate glossary entry.)

Which of these actually helps depends on where the person’s own bottleneck is. Prescribed treatment exists partly for that reason: a supervised dose lets a clinician find the right intervention at the right point on the chain, tune it for the individual, and monitor how it interacts with the rest of the person’s life. Someone reaching for a coffee, a cigarette, or a glass of wine is engaging with the same neurochemistry without the dose control, the pharmacokinetic tuning, or the clinical oversight. Whether that turns into a problem depends on how well it’s actually working.

Beyond dopamine, the general pattern of neurodivergent adults using substances to regulate underlying states has been documented across the population.

In autistic adults, substance use is more often driven by managing anxiety, sensory overload, and low mood than by social reasons⁷. Adults with ADHD show higher use of most stimulant substances than non-ADHD peers¹⁵. The self-medication hypothesis, from Khantzian, proposes that people reach for specific substances to regulate specific distress states rather than for random pleasure-seeking. His hypothesis has held up across decades of subsequent research¹².

Common substances neurodivergent adults reach for

Caffeine

Caffeine crosses the blood-brain barrier easily and blocks adenosine receptors. Adenosine builds up while you’re awake and produces the feeling of gradually mounting sleepiness. Blocking it raises alertness and indirectly raises dopamine availability in the frontocortical and striatal circuits most affected in ADHD⁴.

Animal models show caffeine can normalise elevated dopamine transporter density and improve attention in ADHD brains⁴. If the coffee habit that started at university has persisted through your twenties and thirties without ever quite feeling like a habit but more like a necessity, this is what it has been doing.

Whether it’s actually helping is a separate question from whether it makes intuitive sense. A large survey of adults with ADHD found caffeine use associated with caffeine use disorder and lower overall well-being, rather than with sustained symptom relief⁵. In practice, coffee tends to take the edge off a regulatory need without addressing what’s driving it, and long-term reliance can end up adding a caffeine dependence problem alongside the original one.

Alcohol & the neurodivergent brain

Alcohol is one of the most consistently misused substances in both autistic and ADHD populations. It can slow down baseline arousal. And for nervous systems that run at higher baseline arousal (common in both ADHD and autism), that relief that evening cup of wine brings can feel like the first moment of silence that day.

Alcohol enhances GABA, the brain’s main inhibitory neurotransmitter, and suppresses glutamate, the main excitatory one. The overall effect is a slowing of neural signalling, which produces the familiar calming feeling and can temporarily reduce social anxiety and sensory overload.

Large surveys of autistic adults document high rates of co-occurring anxiety and depression, with alcohol frequently used for self-management⁶⁷. Someone who has been masking through a day of meetings, or who has just left a social event they wanted to be at but found overwhelming, is often reaching for a drink for its quieting effect. The drink is doing regulatory work that would otherwise be done by adaptations and accommodations. Which is completely understandable if the person doesn’t even have a framework for neurodivergent adaptations because they are unaware of their autism. No wonder that the only working tool they have is a few glasses of wine, which they noticed were necessary to not crash out after an event.

The problem with alcohol as regulation is that the GABA-boosting mechanism providing the relief also builds tolerance and rebound anxiety. Kronenberg and colleagues, working with adults in treatment for substance use disorders alongside ADHD or autism, documented how initial relief gave way to escalating use and worsening executive function, with the cycle becoming harder to interrupt the longer it ran¹⁴.

Nicotine & ADHD

Nicotine acts as an indirect dopamine agonist, promoting release in the ventral tegmental area and prefrontal cortex, which are some of the same circuits methylphenidate (Ritalin, Concerta) acts on⁸. Levin and colleagues, in 1996, gave a low-dose transdermal nicotine patch to eleven non-smoking adults with ADHD and found significant improvements in clinician-rated global functioning and continuous performance reaction time. Because the participants weren’t smokers, the improvement couldn’t be attributed to relief from withdrawal, which meant it was a direct pharmacological effect⁸. The trial was small, but the finding was interesting nonetheless.

A follow-up study in non-smoking adolescents with ADHD found that nicotine improved stop-signal reaction time, a measure of behavioural inhibition, in a small sample where methylphenidate had no effect on the same task⁹. Potter and Newhouse made sure nobody could take their findings as a shining endorsement for smoking. Positive cognitive effects that continue with chronic exposure may contribute to the vulnerability adolescents with ADHD have to smoking persistence. Nicotine provides the relief, but it also builds the trap and creates the harmful habit.

For many adults, smoking was just always a part of their lives. Maybe you took up smoking as a teenager and never quite stopped, or picked it up again during a hard stretch at work. Your brain could be getting substantive regulatory work from the nicotine. However, prescribed stimulants target the same regulatory need with better dose control and considerably less dependence risk.

Earlier clinical work found notably low smoking rates in autistic adults, with a Swedish sample from 2003 putting the figure at 12.6% versus 19% in the general population. More recent US convenience samples have found rates closer to or exceeding general population figures¹¹. The likely explanation is co-occurrence. Autistic adults without ADHD, anxiety, or depression are less likely to be reaching for nicotine, whereas autistic adults with those co-occurrences show the same pattern as their ADHD peers.

Herbal tea

Lemon balm (Melissa officinalis) contains compounds that inhibit GABA transaminase, the enzyme that breaks down GABA. The most active of these is rosmarinic acid. The overall effect is that more GABA stays available for longer, and small clinical trials show modest reductions in mild anxiety and improvements in sleep quality¹²¹³. If a cup of chamomile or lemon balm tea in the evening tends to correspond with better sleep, that’s a neurochemical event with a mechanism behind it, gentler than the other substances in this section but the same kind of thing.

Chemical self-regulation is a routine feature of ordinary life. If you have a bedtime tea routine that helps you wind down, you’re already interacting with your own GABA system through diet, and the mechanism is in the same category as everything else in this section.

Substances can work – until they don’t

When someone reaches for a substance that shifts their neurochemistry, they often choose the best available option at the time. It might have been the only option, given how many neurodivergent adults spend years or decades without a framework for what exactly they’re regulating.

The catch with self-medication is that the mechanisms that make certain substances feel essential are also the mechanisms that build dependence, tolerance, and escalation over time. What starts as effective regulation in a nineteen-year-old often stops being sustainable by the mid-thirties, and it is due to how these substances act on the nervous system.

ADHD and addiction

Let’s take nicotine, for example. Ilbegi and colleagues followed a Dutch cohort from childhood into young adulthood. They found that 24.2% of adults with persistent ADHD met criteria for nicotine dependence, compared with 4.3% of controls in the same cohort, a nearly six-fold difference in prevalence¹⁰. The receptor system nicotine acts on is doing more regulatory work in an ADHD brain than in a brain in the control group, which makes the initial exposure more rewarding and the dependence quicker to establish. Looking at this as a bad habit due to insufficient willpower misidentifies its source. Effective self-regulation slowly becomes a dependence that runs alongside the original need it was trying to meet.

Kronenberg’s qualitative work with adults in treatment for substance use disorders and co-occurring ADHD or autism traced how initial relief steadily gave way to escalating use, deteriorating executive function, and a shrinking behavioural repertoire, with the substance moving from an occasional tool to something the person’s day was organised around¹⁴. Falling into that pattern is what happens when the regulatory strategy relies on a dose-dependent, tolerance-building substance without clinical oversight.

The other issue with harmful self-medication is what the chosen substance leaves untouched.

• Caffeine takes the edge off a dopamine deficit, but doesn’t address what’s causing it.
• Alcohol quiets sensory overload but doesn’t change the sensory environment producing it, and continues to be needed as long as the overload keeps arriving.
• Nicotine provides short-term attention regulation while disrupting sleep, which then undermines attention over the medium term.

The thing you need to ponder is whether the current arrangement is the one that’s actually helping, and whether you are willing to keep paying the cost of maintaining it. If you need help in deciding this, please talk to a professional.

ADHD medication actually reduces addiction risk

Much of the reluctance to consider medication comes from the fear that starting a prescribed stimulant means trading one form of dependence for another. Studies actually found evidence for the opposite. Supervised treatment is associated with less substance-related harm than unsupervised self-medication.

Chang and colleagues analysed Swedish national registry data for nearly 40,000 adults with ADHD across four years and found that being on ADHD medication was associated with a 31% reduction in substance abuse rates, after controlling for prior substance abuse, psychiatric history, and criminal convictions¹⁶. The longer someone stayed on medication, the lower the rate: each additional year of treatment was associated with a further 13% reduction.

Quinn and colleagues, working with US commercial health insurance claims for nearly three million ADHD patients, ran within-individual comparisons that use each patient as their own control, by comparing months when the same person was on medication with months when they weren’t. Men on medication had 35% lower odds of substance-related emergency events during medicated months; women had 31% lower odds. Two years after medicated periods, men still had 19% lower odds, and women 14% lower¹⁷.

There are two reasons stimulant medication can have a protective effect.

1. Prescribed stimulants reduce the underlying regulatory symptoms (impulsivity, attention volatility, mood variability) that drive substance use vulnerability in the first place.
2. They also reduce the day-to-day exposure to substance-using contexts that comes with unmedicated ADHD, from the friend groups formed around shared coping to the environments where drinking or smoking is the default. Someone whose regulatory needs are being met more head-on by a supervised, dose-controlled intervention has less incentive to reach for the unsupervised alternative.

But don’t worry if you don’t have a time machine to advocate for your teen self. It is still not too late for late-identified adults starting treatment in their thirties, forties, or fifties; the protective effect still applies. The trajectory of change is likely different than it would be for someone who started at fourteen, but the benefits are still there.