# ADHD Is Not Simply “Low Dopamine”: What Studying the Biology and My Own Genetics Has Taught Me

> I have been looking into ADHD biology for a very long time, particularly the genetic and molecular side of it, and because I have ADHD myself and have been formally diagnosed with it, this subject has always been personal to me as much as

- **Author:** Mohammed Attallah (BiomeLogic)
- **Published:** 2026-09-11
- **Category:** Host Capacity Model
- **Tags:** adhd, serotonin, dopamine, cortisol
- **Canonical URL:** https://biomelogic.net/articles/adhd-is-not-a-dopamine-deficiency
- **License:** CC BY-NC 4.0 — please cite "Mohammed Attallah, BiomeLogic" with link to https://biomelogic.net/articles/adhd-is-not-a-dopamine-deficiency.

---

I have been looking into ADHD biology for a very long time, particularly the genetic and molecular side of it, and because I have ADHD myself and have been formally diagnosed with it, this subject has always been personal to me as much as scientific. I know what it is like to live with one of the central contradictions of ADHD: sometimes attention can feel almost impossible to stabilize, while at other times it can become so intense that hours disappear. Something novel, complicated, urgent, emotionally meaningful or immediately rewarding can hold attention almost effortlessly, while a repetitive task with a distant reward can require an enormous amount of effort just to keep the mind in the same place. That observation alone has always made the phrase “attention deficit” feel incomplete to me. The capacity for attention clearly exists. The deeper question is how attention is allocated, stabilized, interrupted and influenced by reward, stress, arousal, novelty and context.

For the same reason, I have never found the explanation that ADHD is simply a condition of “low dopamine” very satisfying. Dopamine absolutely matters. It participates in motivation, reinforcement learning, reward prediction, behavioral selection and executive function, but once dopamine is placed back into the biological systems surrounding it, the picture becomes much larger. Norepinephrine, intracellular cAMP signaling, HCN channels, glutamate, calcium, GABAergic inhibition, synaptic vesicle release, neuronal adhesion, cytoskeletal architecture, stress signaling, circadian regulation and developmental gene programs all become relevant. The more complete the biology becomes, the harder it is to reduce ADHD to the concentration of one neurotransmitter.

One of the first things I learned from looking more deeply into my own genetics is that ADHD genetics cannot be reduced to a list of “good” and “bad” SNPs. ADHD is highly polygenic. There is no single ADHD gene, and there is no one common variant that explains why somebody develops ADHD. Thousands of common variants can contribute very small effects, while rarer variants, copy-number changes and structural variants may contribute larger effects in a smaller number of people. Some variants influence susceptibility statistically, some alter neurotransmitter physiology, some modify stress response or intracellular signaling, and others influence how neurons grow, organize, form synapses, position ion channels and respond to experience. The useful way to think about ADHD genetics is therefore not as a list of isolated SNPs but as a collection of biological systems interacting across development.

Even the basic catecholamine pathway shows why a simple dopamine model is inadequate. Tyrosine enters catecholamine synthesis through tyrosine hydroxylase, producing L-DOPA. Aromatic L-amino-acid decarboxylase converts L-DOPA into dopamine. Dopamine can then be converted by dopamine beta-hydroxylase into norepinephrine. Catecholamines are packaged into vesicles through VMAT2, released into the synapse when calcium triggers vesicle fusion, bind different receptor subtypes, are removed through transporters such as DAT and NET, and are metabolized through enzymes including COMT and monoamine oxidases. Every one of these stages is a distinct biological control point. Altered synthesis is not the same as altered reuptake. Altered receptor density is not the same as altered receptor affinity. Altered extracellular dopamine is not the same as an altered intracellular response to dopamine. A neuron can have a normal amount of neurotransmitter stored inside it and still communicate differently if vesicle release, receptor coupling or downstream signaling differs.

One of the most interesting findings from my own data involved ANKK1 and DRD2 because this is an area where it is very easy to create an appealing story. The famous ANKK1 variant rs1800497, historically called Taq1A, has been studied extensively because the A1 allele has been associated with differences in striatal D2 receptor availability. Because of that, it has often been discussed in relation to reward sensitivity, addiction, impulsivity and what has sometimes been called reward deficiency. It would have been easy for me to take my ADHD diagnosis, my own relationship with stimulation and reward, and conclude that perhaps I carry the classic A1 allele and therefore have reduced D2 receptor availability.

But I do not. My ANKK1 rs1800497 genotype is GG, which means I do not carry the classic A1 allele. That matters because it removes one of the simplest versions of the argument that my ADHD is primarily an inherited ANKK1-driven reduction in D2 receptor availability. It does not mean D2 signaling is irrelevant, and it does not mean my reward circuitry can be declared “normal” from one SNP. It simply means that one very specific and commonly cited mechanism does not fit my genotype.

I then went deeper into DRD2 itself. Two variants, rs1076560 and rs2283265, have been studied because their minor alleles can influence alternative splicing between D2 receptor isoforms. My genotype is CC at rs1076560 and CC at rs2283265, so I do not carry those particular minor splicing-associated alleles. I am heterozygous at DRD2 rs6277, with a GA genotype, and that variant has been investigated in relation to receptor affinity and apparent D2 availability. Taken together, this looks much more like subtle modulation of D2 biology than a dramatic inherited D2 deficiency.

That distinction is important because dopamine receptor biology is not one-dimensional. D2-family receptors participate in striatal reinforcement learning, behavioral selection and feedback regulation of dopaminergic signaling. D2 and D3 autoreceptors can function as part of the braking machinery limiting dopamine-neuron activity and dopamine release. DRD3 has relatively strong expression in limbic and reward-related regions, while DRD4 and DRD5 have their own distributions and signaling properties. Even before we leave the dopamine system, we are already dealing with multiple receptor subtypes, different anatomical regions, different cell types and different intracellular effects.

My COMT genotype adds another layer. I am COMT rs4680 GG, or Val/Val. COMT metabolizes catecholamines including dopamine and norepinephrine, and the Val158 form generally has greater enzymatic activity than the Met158 form. The important part, however, is anatomical context. Dopamine clearance is not identical throughout the brain. DAT plays a major role in the striatum, whereas parts of the prefrontal cortex have relatively less DAT activity and rely more heavily on NET and enzymatic mechanisms including COMT. COMT Val/Val therefore does not mean “my entire brain has low dopamine.” It suggests that one component of prefrontal catecholamine clearance may be relatively more active.

That becomes important because prefrontal cortical function appears to depend on a relatively narrow optimal catecholamine range. Too little dopamine or norepinephrine can weaken working memory, behavioral inhibition and executive control, but too much can impair those same functions. The relationship is better described by an inverted-U than a simple “more is better” model. This is one reason stimulant pharmacology should not be reduced to “raising dopamine.” The useful effect depends on where the system begins and whether catecholamine signaling moves particular cortical and striatal circuits toward a more optimal operating state.

Norepinephrine becomes especially important here because one of the most elegant models of prefrontal function involves the alpha-2A adrenergic receptor, cAMP and HCN channels. The prefrontal cortex contains recurrent neuronal networks that need to keep firing after the original stimulus has disappeared. That persistent activity helps the brain maintain a goal, rule or piece of information in working memory. If I decide that I need to finish a report before doing something more interesting, the representation of “finish the report” has to remain active strongly enough to compete with more immediately rewarding alternatives.

Norepinephrine acting at postsynaptic alpha-2A receptors can inhibit cAMP through Gi-coupled signaling. Reduced cAMP decreases HCN-channel opening in dendritic spines, helping strengthen functional connectivity within those recurrent networks. The representation becomes more stable. My own genotype includes ADRA2A rs1800544 GC. I do not consider this an “ADHD mutation,” but I find it interesting because it sits directly inside a pathway relevant to prefrontal network stability.

The opposite state is equally important. If cAMP rises excessively in the wrong microdomain, HCN channels can open and weaken those recurrent connections. Now the task representation becomes less stable, and a more salient competitor can win. This gives us a much better framework for understanding one of the classic paradoxes of ADHD. The same person can sometimes focus intensely for hours on something fascinating and then struggle for fifteen minutes with something repetitive. The capacity for attention is not absent. The problem may involve which network receives priority, how strongly it is stabilized and how long it remains dominant.

Dopamine D1 signaling interacts with this cortical system as well. Moderate D1 activation can improve signal-to-noise by suppressing irrelevant activity, while excessive D1 stimulation can become destabilizing. This is where stress becomes especially important. Acute stress can sharply increase dopamine and norepinephrine, pushing prefrontal networks beyond their optimal range. A person can therefore function reasonably well in one state and lose executive control in another even though the underlying anatomy has not changed.

This is where another one of my genotypes becomes interesting: FKBP5 rs1360780 TT. FKBP5 is not an ADHD gene in a simple diagnostic sense. It regulates glucocorticoid-receptor signaling and participates in HPA-axis feedback. What interests me is not the idea that FKBP5 “caused” my ADHD, but that stress-response biology may modify how strongly an underlying ADHD phenotype is expressed. I find the distinction between trait architecture and state regulation much more useful. A developmental network may establish baseline vulnerability, while sleep deprivation, chronic stress, cortisol, anxiety, circadian disruption or environmental overload determine how strongly that vulnerability appears on a particular day.

I also examined DBH because dopamine beta-hydroxylase converts dopamine into norepinephrine. If somebody had substantially reduced DBH activity, that could theoretically shift catecholamine balance. My genotype at the major promoter variant DBH rs1611115 is CC rather than the low-activity TT configuration. So once again, the data move me away from another simple explanation. My ADHD does not appear to be easily explained by an inherited failure to convert dopamine into norepinephrine.

The next level is what happens after a neurotransmitter binds its receptor. Receptor activation is only the beginning of a much larger intracellular signaling cascade. This is why I have been interested in the PDE4 family, including PDE4D. I would not describe PDE4D as one of my established ADHD genes because I do not currently have a specific PDE4D variant with enough evidence to justify that claim, but the pathway is extremely important. PDE4 enzymes break down cAMP and therefore influence the duration and amplitude of intracellular cyclic-nucleotide signaling.

Dopamine or norepinephrine reaches a receptor, the receptor changes a G protein, the G protein alters adenylyl cyclase, cAMP rises or falls, PKA changes phosphorylation states, ion channels change their opening probability, receptor trafficking can change, and transcription factors such as CREB may eventually influence gene expression. Phosphodiesterases determine how quickly some of that signal is terminated. Two brains could therefore receive almost identical extracellular dopamine signals and still produce very different outputs because everything downstream of the receptor differs. One may have stronger receptor coupling, another may degrade cAMP more rapidly, another may express a different complement of ion channels, and another may differ in receptor trafficking or phosphatase activity.

This is why I increasingly think the better question is not simply, “How much dopamine does somebody have?” but, “What does the network do when dopamine arrives?”

Then there is the machinery controlling neurotransmitter release itself. My genotype includes SNAP25 rs3746544 TT. SNAP25 is part of the SNARE complex involved in synaptic vesicle fusion. Neurotransmitters are stored inside vesicles. When an action potential reaches the presynaptic terminal, voltage-gated calcium channels open, calcium enters and proteins including synaptotagmin, syntaxin, VAMP and SNAP25 help the vesicle fuse with the membrane and release its contents.

This introduces time into the ADHD model. Neural computation does not depend only on the concentration of a neurotransmitter. Release probability matters. Presynaptic calcium matters. Synchronization between neuronal populations matters. The exact timing of excitatory and inhibitory input matters. Milliseconds can determine whether a neuron crosses threshold and whether a network stabilizes or collapses. A person can therefore have normal neurotransmitter synthesis yet still show different network behavior if the timing or probability of release differs. Certain SNAP25 variants have shown modest associations with ADHD in older candidate-gene research, but those effects are small and inconsistent, so I do not consider rs3746544 an ADHD mutation. I consider the biology of the pathway more important than the label placed on one SNP.

The deeper I went into ADHD, the more important glutamate became. Dopamine is useful for understanding salience, motivation, reward prediction and reinforcement, but dopamine alone cannot explain how experience becomes lasting learning. Glutamatergic signaling is fundamental to activity-dependent synaptic plasticity. My raw data include GRIN2B rs1805502 AG, GRM3 rs6465084 AA and GABRA2 rs279858 TC. I am deliberately not calling these causal ADHD variants. The association evidence for those specific SNPs is not strong enough. Their importance to me is that they sit inside systems fundamental to learning, excitation, inhibition and network regulation.

NMDA receptors are particularly interesting because they function partly as coincidence detectors. Glutamate must bind, but the postsynaptic membrane also has to be sufficiently depolarized to relieve the magnesium block in the receptor pore. When those conditions occur together, calcium can enter the cell. Calcium then becomes an intracellular information signal, activating pathways such as CaMKII, altering AMPA-receptor trafficking, modifying dendritic-spine structure, influencing local protein synthesis and changing transcription. A stimulus lasting seconds can therefore alter synaptic strength for much longer.

This is where dopamine and glutamate become inseparable. Dopamine can help signal that something was important, unexpected or rewarding, while glutamate-dependent plasticity helps modify the synapses that participated in that event. In simplified language, dopamine can say, “That mattered,” while glutamatergic plasticity helps determine which connections should become stronger because it mattered.

This framework makes delay discounting much easier to understand. A future reward does not physically exist in the present. The brain has to represent it internally. The prefrontal cortex has to keep that future outcome active strongly enough to guide behavior now. An immediate reward has a computational advantage because it is present, salient and reinforced immediately. If the representation of the delayed goal loses stability more easily, behavior can repeatedly shift toward whatever is available now even when the person intellectually values the future outcome more.

From the outside that can look like laziness or lack of discipline. Biologically it may reflect competition between networks operating on different timescales. A person can fully understand what needs to be done and still fail to keep the future consequence dominant enough to guide behavior. Knowledge and executive implementation are not the same neural process.

Glutamate also cannot be discussed without GABA. Cortical function depends on precisely coordinated excitation and inhibition. GABAergic interneurons shape oscillations, timing, gain, synchronization and the boundaries of neuronal ensembles. The relevant problem is therefore rarely “too much glutamate” or “too little GABA” throughout the entire brain. What matters is where the signal occurs, in which neuronal population, at what time and relative to what competing input. Brain function is spatial, temporal and network-dependent.

BDNF provides another example of why biological importance and genetic prediction should not be confused. My common BDNF rs6265 genotype is CC, corresponding to Val/Val. BDNF is enormously important to neuronal development, dendritic growth and activity-dependent plasticity, but the common Val66Met polymorphism has not become a strong or reliable ADHD susceptibility marker. A gene can be central to nervous-system biology without one common SNP strongly predicting a diagnosis.

The same is true of CACNA1C. My rs1006737 genotype is GG. CACNA1C is important to voltage-dependent calcium signaling and psychiatric neurobiology broadly, but its biological importance should not be confused with an ADHD-specific explanation. This distinction matters because older candidate-gene discussions often mixed mechanistic plausibility with actual disease association.

Genes such as DRD4, SLC6A3, SLC6A2, MAOA, MAOB, TH, CHRNA4, SLC6A4, TPH2 and multiple serotonin receptors all participate in systems relevant to attention, motivation, emotional regulation or impulsivity. That makes them interesting. It does not automatically mean that every common SNP within them is an ADHD risk variant.

My own data include SLC6A2 rs5569 GA within the norepinephrine transporter gene and SLC6A3 rs27072 CC within the dopamine transporter gene. But even those findings reveal another technical limitation. Several historically famous ADHD variants are repeat polymorphisms rather than ordinary SNPs. The DAT1/SLC6A3 VNTR, DRD4 exon-3 VNTR and MAOA promoter VNTR cannot be reliably reconstructed from a few nearby SNPs in a simplified genotype file. So looking at a raw compatibility file does not mean every relevant type of genetic variation has been characterized.

Then there is ADGRL3, historically called LPHN3, which interests me because it moves the discussion away from neurotransmitter quantity and toward neuronal organization. ADGRL3 is an adhesion GPCR involved in neuronal and synaptic organization. Once neuronal adhesion genes enter the model, ADHD begins to look less like a mature brain that simply lacks enough dopamine and more like something involving how networks were assembled during development.

Modern ADHD genetics has pushed the field further in that direction. Genes and loci involving SORCS3, MEF2C, PTPRF and related pathways point toward neuronal development, synaptic organization and activity-dependent plasticity. More recent rare-variant work involving MAP1A, ANK2 and ANO8 shifts the model even further toward neuronal architecture itself. MAP1A participates in microtubule and cytoskeletal organization. ANK2 encodes ankyrin-B, which helps organize membrane proteins, ion channels and neuronal structure. ANO8 brings membrane and calcium-related signaling into the picture.

This changes the conceptual model substantially. Instead of imagining a normally constructed brain that simply happens to be short on dopamine, imagine a developing nervous system in which thousands of genetic differences subtly influence neuronal migration, axonal growth, dendritic structure, synapse formation, receptor trafficking, membrane organization, ion-channel placement, cytoskeletal architecture, calcium signaling and the way large cortical and striatal networks become connected.

ADHD may therefore begin partly with how the network was built, long before a child is old enough to describe what attention feels like.

That does not mean everybody with ADHD carries rare damaging variants in MAP1A, ANK2 or ANO8. The importance of those findings is that they reveal the types of biological systems capable of contributing to the phenotype. Different people may arrive at the same diagnostic category through partially different molecular routes.

The same principle applies to common-variant genetics. My raw file contains genotypes at two lead variants identified in modern ADHD genome-wide work: rs10875612 TT and rs704061 TC. Their individual effects are small, and that is exactly what modern polygenic genetics predicts. Most common ADHD-associated alleles do not produce dramatic changes in risk. Many shift probability only slightly. Their importance appears when thousands of variants act together across development.

This is why I no longer think statements such as “40% of my dopamine genes are bad” or “60% of my glutamate genes are affected” are scientifically meaningful. That sounds quantitative, but it is not a valid measure of genetic liability. One panel may contain many more SNPs from one pathway than another. Several variants may travel together because they are in linkage disequilibrium. One SNP may have a tiny statistical effect while another changes protein function substantially. One may be extremely common in the general population, while another may not replicate consistently at all. Counting all of them as equal red flags creates false precision.

A genuine ADHD polygenic score is a completely different analysis. It requires very large GWAS datasets, thousands to millions of variants, effect-size weighting, linkage-disequilibrium modeling, good genotype coverage or imputation and ancestry-matched reference populations. Even then, a polygenic score estimates probability. It does not tell an individual exactly why they have ADHD.

The more useful distinction is between genetic risk architecture and functional modifiers. COMT Val/Val may influence prefrontal catecholamine physiology without being a major ADHD risk allele. FKBP5 TT may influence stress responsivity without being an ADHD gene. ADRA2A may modify prefrontal noradrenergic signaling without determining whether somebody develops ADHD. SNAP25 may sit inside highly relevant synaptic machinery while one common SNP produces only a small effect. This layered interpretation is much more realistic than trying to force every SNP into a causal category.

At one level there is developmental architecture: common variants, rare variants, regulatory changes, structural variants and gene-expression programs influencing how neurons are produced, migrate, extend axons and form synapses. At another level there is synaptic architecture: adhesion molecules, scaffolding proteins, cytoskeleton, vesicle-release machinery, receptors, ion channels and excitatory-inhibitory balance. Then there is neuromodulation: dopamine, norepinephrine, serotonin, acetylcholine and other systems adjusting network gain, flexibility, salience and motivational value. Then there is intracellular signal processing: cAMP, PKA, calcium, CaMKII, ERK/MAPK, phosphodiesterases, phosphatases, receptor trafficking and transcription factors. Finally, there is state regulation: sleep, stress, cortisol, circadian phase, hormonal state and arousal continuously shifting how the same circuitry behaves.

Experience-dependent plasticity interacts with all of those layers, and this is where FOSB and ΔFosB become particularly interesting. ΔFosB is not an inherited ADHD mutation. It is a transcription factor produced from the FOSB gene, and what makes it unusual is its stability. Many immediate-early transcription factors rise after stimulation and disappear rapidly. ΔFosB can accumulate with repeated stimulation and persist much longer in reward-related circuitry.

That means repeated reward is not simply a sequence of isolated dopamine spikes that appear and disappear without leaving any biological consequence. Repeated neural activation can eventually alter transcription. A transient signal can become a more persistent cellular adaptation.

This provides a bridge between inherited architecture and experience. Someone may begin life with a particular executive-control system, reward sensitivity, stress response and tendency toward delay discounting. That nervous system is then trained every day by what it repeatedly experiences. Drugs, gambling, pornography, gaming, highly palatable foods, compulsive shopping, social media and constant novelty are not biologically identical, but they illustrate the broader principle that repeated high-salience reinforcement can reshape the circuitry repeatedly recruited by that experience.

The nervous system adapts to what it repeatedly does.

This is why I think the relationship between ADHD and addiction is much more sophisticated than saying people with ADHD are “chasing dopamine.” The more useful questions are how strongly immediate rewards are valued, how steeply future rewards are discounted, how stable a long-term goal remains when an immediate reward appears, how quickly the striatum updates reward value, how strongly novelty captures attention and how repeated reinforcement changes those circuits over time.

The same framework helps explain impulsivity. Impulsivity does not necessarily mean ignorance of consequences. A person can fully understand tomorrow’s consequence and still struggle to keep that future consequence computationally dominant when an immediate alternative appears. Understanding a goal and maintaining that goal strongly enough to control behavior are different neural functions.

Circadian biology adds another layer. Genes such as CLOCK and ARNTL are interesting because the circadian system regulates far more than sleep. Circadian transcription influences endocrine signaling, metabolism, arousal, catecholamine physiology and cortical function. Sleep deprivation alone can impair prefrontal function in somebody without ADHD. If someone already has a vulnerable executive network, chronic circadian disruption can amplify the phenotype substantially.

The cholinergic and serotonergic systems also matter. CHRNA4 and related nicotinic receptor genes are relevant to attentional gating and sensory selection. SLC6A4, TPH2 and HTR-family receptors remind us that impulsivity, punishment sensitivity, patience and emotional regulation do not belong exclusively to dopamine. MAOA and MAOB influence monoamine degradation. TH regulates a rate-limiting step in catecholamine synthesis. DBH determines the conversion of dopamine into norepinephrine. Calcium-channel genes regulate activity-dependent calcium entry. These systems intersect continuously rather than functioning as isolated pathways.

This may be one reason two people can both meet criteria for ADHD and yet look very different biologically and behaviorally. One person may be dominated by reward sensitivity and delay discounting. Another may have profound working-memory instability. Another may show stronger emotional impulsivity. Another may be heavily affected by sleep and circadian disruption. Another may function relatively well until stress becomes high. Another may have broader differences in cortical maturation or sensory processing. Most people are probably mixtures rather than clean categories.

That is why I think the future of ADHD biology will eventually move beyond simply asking whether somebody “has ADHD” and toward a more informative question: what biological architecture produced that person’s ADHD phenotype?

Looking into my own genetics has made me more cautious rather than more deterministic. I do not see one gene that explains my ADHD. I do not see convincing evidence that my ADHD can be reduced to an ANKK1-driven D2 receptor deficiency. In fact, my ANKK1 genotype argues against the classic version of that explanation. I do not carry the particular DRD2 minor alleles at rs1076560 and rs2283265 that would support the specific splicing mechanism I examined, and my major DBH promoter genotype does not support an obvious low-DBH model.

At the same time, I carry COMT Val/Val, which makes prefrontal catecholamine clearance physiologically interesting. I carry ADRA2A rs1800544 GC, which sits within the alpha-2A/cAMP regulatory system. I carry FKBP5 rs1360780 TT, which makes stress-dependent modulation interesting. I carry SNAP25 rs3746544 TT, which connects to synaptic-release machinery even though the evidence for that particular SNP is modest. I also carry variants in glutamatergic and inhibitory pathways and genotypes at modern ADHD GWAS loci, but none of these individually explains the diagnosis.

The most defensible model I currently see is therefore not “my brain cannot make enough dopamine.” It is a broader polygenic neurodevelopmental architecture influencing the construction of executive, attentional and reward networks. Dopamine and norepinephrine modulate the gain and stability of those networks. Alpha-2A signaling and cAMP influence whether prefrontal representations remain stable. Glutamate and calcium determine how experience alters synaptic strength. Presynaptic machinery determines the timing of neurotransmitter release. Stress biology can move the system away from its optimal operating range. Circadian state can change cortical efficiency. Repeated reward exposure then trains the circuitry over time.

That model explains things a simple dopamine-deficiency model never could. It can explain how the same person can be extraordinarily focused under one condition and profoundly distractible under another, why urgency can suddenly unlock performance, why novelty can overpower intention, why a future goal can feel intellectually important and still lose the behavioral competition against something immediately rewarding, why stress can make executive control deteriorate and why environmental changes can alter performance without changing the underlying genetics.

It also explains why medications that alter catecholamine signaling can help even though ADHD is not simply a catecholamine deficiency. A treatment does not necessarily have to replace something that is missing. It can change the operating state of a network.

I have ADHD. I have been diagnosed with ADHD, and living with it has made me very aware that it cannot be described simply as an inability to concentrate. My genetics do not reduce that experience to one mutation, one neurotransmitter or one pathway. If anything, studying my own genome has shown me how dangerous oversimplification can be.

The deeper I look into ADHD, the more it appears to be a neurodevelopmental disorder of network regulation, neural gain, temporal stability, working-memory maintenance, reinforcement learning, reward valuation, executive control and plasticity. Dopamine is part of that system, but so are norepinephrine, glutamate, GABA, calcium, cAMP, synaptic release, developmental wiring, stress, circadian biology and experience.

What has been most useful to me in examining my own genetics is not finding one perfect explanation. It has been learning which explanations I have to abandon when the data do not support them. The ANKK1 finding is a good example. If I wanted a convenient low-D2 reward-deficiency story, I could have constructed one. My actual genotype forced me to let that version go.

That is what data should do. It should not simply give us a more sophisticated way to confirm what we already believe. Sometimes its greatest value is forcing us to rebuild the model.

Right now, the model that makes the most sense to me is not a brain that simply lacks dopamine. It is a developing nervous system whose structure, signaling, gain, timing, learning and regulation differ through thousands of interacting biological influences.

That is where ADHD becomes much more biologically interesting.

---

*Source of record: https://biomelogic.net/articles/adhd-is-not-a-dopamine-deficiency. This plain-text endpoint is provided for AI retrieval and citation systems (ChatGPT, Perplexity, Claude, Gemini, Bing Copilot).*