If you’re taking stimulants daily and still can’t focus for more than 15 minutes, you’re probably depleting cofactors faster than you’re replacing them.
The status quo is terrible: supplement guides are just trying to sell you things or based on poor (sometimes straight up fictious) information; governmental recommendations are based on population level advice, meaning that it might stop you getting scurvy, but it won’t do much more.
Foreword
A multivitamin may seem like a good idea, but profit is the driving factor for manufacturing companies, and the industry is largely unregulated compared to pharmaceuticals. Generally, they aim to fulfil the governmental recommendations whilst cutting as much cost, taking advantage of the average consumer being less knowledgeable than those who buy individual supplements.
Unfortunately, they amalgamate compounds chosen for shelf stability rather than effectiveness, doses supposedly aimed at not harming entire populations — whilst doing exactly that. Synthetic B6 causing neuropathy, toxic levels of vitamin A in populations with adequate intake, etc. They’re largely incorrect in both form and amount.
This review aims to largely solve the problem of vitamin, mineral, and nutrient with as little intervention as possible. Whilst it was written with primarily ADHD in mind, people without it can still benefit from both the architecture and methodology described. An alternative title could have been ‘The Pareto Principle Supplement Stack’; but that would be understating the time saved, this gives you 90% of the relevant information with 10% of the complexity, without hiding justifications or methodology.
It’s written as a series of interlinked hedges derived from each supplement by asking: What does it do? What’s the benefit? What’s the risk? How can I supplement this blind? It’s designed for fault tolerance, not optimisation, whilst still being more efficient than most generic stacks out there.
Disclaimer: This is medical advice. If you can read it, then you’re capable of making your own decisions.
It primarily attempts to remain directly actionable. The supplements are split into two main groups: morning and evening, based on the intended time of dosing. Headings serve as a skimmable overview of the entire stack. Those marked as ‘as needed’ should be read in full to understand both when and how to take them, otherwise they can be safely ignored entirely.
A few prescription medications are included as useful information is particularly scattered; doctors who prescribe them generally do not do much in the way of actually telling you what they are, what they do, and how they interact with the everything else.
Morning
Creatine Monohydrate — 5 g
Creatine’s mechanism is well understood: it acts as a buffer for ATP, the cell’s energy currency, in the form of phosphocreatine maintaining energy availability during periods of high demand (Allen, 2012).
Supplementation enhances short-term memory and abstract reasoning (Hammett, 2010), with the strongest effects when phosphocreatine reserves are strained. At baseline, well-rested non-vegetarians often show little change (Rawson, 2008), whilst vegetarians and vegans, who generally have lower stores, benefit more (Rae, 2003). Under stress or sleep loss, however, benefits extend to everyone: supplementation reduced mental fatigue (Watanabe, 2002), improved mood and physical fatigue during sleep deprivation (McMorris, 2006), and enhanced executive and working memory with greater sleep loss (McMorris, 2007). Gains in verbal, spatial, and long-term memory can emerge within just a week of use (McMorris, 2007). Notably, under similarly sleep-deprived circumstances, even single-dose creatine improved cognition (Gordji-Nejad, 2024).
The cognitive benefits extend beyond performance; a large U.S. cohort study (Bakian, 2020) found a significant inverse correlation between dietary creatine intake and depression risk. Kious (2019) propose that depression may involve impaired brain bioenergetics, with early trials suggesting creatine supplementation can restore phosphocreatine buffering and improve depressive symptoms.
While often introduced with a loading phase, this is unnecessary and more likely to cause digestive issues; steady intake of 3–5 g/day achieves the same tissue saturation over time (Hultman, 1996; Buford, 2007; Kreider, 2017). This moderate range is well supported in the literature and avoids potential down-regulation of endogenous creatine synthesis seen with excessive doses (Brosnan & Brosnan, 2007). Adverse effects are generally limited to mild gastrointestinal discomfort at higher intakes (Jäger, 2011). It is typically well tolerated long-term in healthy subjects (Poortmans & Francaux, 1999; Poortmans & Dellalieux, 2000).
Vitamin D3 — 50 µg (2000 IU)
Prevalence of vitamin D deficiency is generally high across the healthy population (Gordon, 2004; Vierucci, 2014). The most well known role of vitamin D is increasing intestinal calcium absorption (Christakos, 2011). This supports bone health by maintaining adequate serum and cellular calcium levels reducing the need to withdraw from skeletal stores (LeBoff, 2022), leading to a lower prevalence of osteoporosis (Voulgaridou, 2023).
Beyond bone health, vitamin D deficiency has been linked to cognitive impairment and immunosuppression among other conditions (Nair & Maseeh, 2012). Various studies found an association of lower vitamin D levels with depression and other psychiatric disorders (Anglin, 2013; Bičíková, 2015; Jeon, 2023). Meta-analyses show some evidence the relation between supplementation and lower risk of depression could be causal (Shaffer, 2014; Jeon, 2023; Lavigne & Gibbons, 2023). And though the link is uncertain (Guzek, 2023), supportive research suggests that, with the associated suicide risk, evidence of improvements to depression (Somoza-Moncada, 2023; Srifuengfung, 2023) is enough to warrant supplementation as adjunctive treatment.
Risk factors for hypovitaminosis D include geographic location. Higher latitudes reduce vitamin D effective UVB (Ladizesky, 1995; Engelsen, 2010; Haggarty, 2013). This is exacerbated in the winter, beginning its decline at ≥N°25, being reduced to almost zero at roughly ≥N°35 leaving most of North America, Europe, Central and North Asia affected (Kimlin, 2007). Additionally, SAD prevalence spikes in these latitudes, adding another reason to correct deficiency (Kerr, 2013; Kerr, 2015).
Recommended sun protection further compounds risk, though this shouldn’t discourage these practices. SPF 30–50 can block over 95% of the UVB radiation needed for vitamin D3 (cholecalciferol) synthesis in the skin (Holick, 2011; Grigalavicius, 2016), and even typical sunscreen under-application can contribute meaningfully to vitamin D deficiency (Neale, 2019; Passeron, 2019).
Given these widespread risk factors, supplementation makes sense for most people. Fortunately, vitamin D has a unique safety profile: Exposure to continued UVB radiation cannot cause toxicity as excess previtamin D3 is photoconverted to biologically inactive forms as a protective mechanism against overdose (Bikle, 2012). There is low risk associated with long term supplementation (Pérez-Castrillon, 2022) and it is generally well tolerated (Albasheer, 2025).
Both the safety profile and pharmacokinetics of vitamin D support regular supplementation. Natural photoconversion prevents sun-induced toxicity, and given the extended half-life (roughly 2–3 months) of 25-Hydroxyvitamin D (Jones, 2008), serum levels may only stabilise after about five to six half-lives which is roughly a year in this case. As a general rule, with each half-life the gap to steady state halves — 50% after one, 75% after two, >95% only after five (Heaney, 2003). Fat tissue provides some buffering but cannot maintain sufficiency across a full winter (Martinaityte, 2017). This means supplementation has to be continuous year-round; whether daily, weekly, or monthly matters less, since the long half-life smooths intake over time (Ilahi, 2008).
The key consideration is maintaining an average intake that meets the updated RDA of 50 µg (2000 IU), as the extended half-life permits dose flexibility unlike vitamins with rapid turnover. And given the long stabilisation time and individual variation in absorption, testing 25(OH)D3 levels after 6–9 months of supplementation can help confirm you’re reaching optimal ranges of 50-75 ng/mL (Pludowski, 2024).
Vitamin K2 (MK-7) — 200 µg
Also playing a central role in calcium metabolism, vitamin K promotes bone strength (Xiao, 2013; Mandatori, 2021), and is linked with a decrease the risk of heart disease though further studies are needed to determine causality (Geleijnse, 2004; Chen, 2024).
Biochemically, activation of two main enzymes is responsible: found in bone, the osteocalcin enzyme is responsible for binding hydroxyapatite — the main component of bone (Hauschka, 1989; Florencio-Silva, 2015); and MGP inhibits vascular calcification (Schurgers, 2013), reducing the risk of decreased blood flow via continued buildup of calcium (Viegas, 2015).
Two primary forms of K2 have been studied, namely MK-4 and MK-7, found in plant and bacterial (or animal) sources (Schwalfenberg, 2017). Research on bioavaibility of different menaquinones show MK-4 does not appreciably increase serum levels unlike MK-7 (Sato, 2012). While MK-4 shows poor serum bioavailability, whether it has meaningful tissue-specific effects independent of serum levels remains unclear.
There is a significant difference in half-lives, with MK-4 showing rapid plasma clearance (on the order of ~1–2 hours), but redistribution into Low-Density Lipoproteins means appreciable amounts remain in circulation for several hours (Schurgers & Vermeer, 2002; Shearer & Newman, 2008). In contrast, MK-7 persists much longer, with a half-life of ~2–3 days, supporting more stable serum levels (Schurgers, 2007). Much like D3, this is useful for practical dosing considerations.
The true required human intake is largely unknown, but some studies suggest that the current adequate intake of 90–120 µg (EFSA, 2017) may be insufficient to gain all the benefits for bone health, though vascular outcomes plateau (Kamao, 2007; Beulens, 2013).
The suggested intake was previously based on biomarkers showing inactive MGP. Using a novel method for estimation of the required intake for maintenance of bone health Tsugawa (2012) inferred that it was 3 times higher than their findings for blood factors, 180-vs-60 µg respectively, suggesting current recommendations may be considered very conservative (Tsugawa & Shiraki, 2020).
No adverse effects were displayed from vitamin K at 10 mg daily in humans; though from limited data, this is supported by the same result for up to 2g in animal models (SCF, 2003).
Inaba (2015) demonstrated that 180 µg/day MK-7 significantly improved osteocalcin γ-carboxylation, suggesting that doses in the ~180–200 µg range are effective for influencing vitamin K-dependent proteins, though clinical outcome data remain limited.
As previously discussed, vitamin D increases calcium absorption; this elevates demand for vitamin K-dependent proteins like osteocalcin and MGP to properly utilise that calcium. And while clinical outcome data for K2 supplementation remain limited, the convergence of several factors supports a pragmatic approach: excellent safety profile, clear biochemical synergy with vitamin D supplementation, biomarker evidence supporting ~180µg doses, and minimal dietary intake in most populations. Given these considerations, 200 µg daily represents a reasonable hedge against functional deficiency.
Vitamin A (Retinyl-acetate) — none
Vitamin A is commonly grouped along with vitamin D3 and K2 for their properties as fat soluble vitamins. This makes logical sense as they all require dietary fat for optimal uptake and share similar transport mechanisms which, taken together, increases absorption.
Generally, toxicity of vitaminosis A is high (Penniston & Tanumihardjo, 2006), with the most harmful effect being liver damage (Minuk, 1988; Barker & Blumsohn, 2003). The margin between having sufficiency and overdose is unusually narrow with even less than the recommended 3000 µg daily upper limit observed to show harm (Arlappa, 2022; Chen, 2023).
There is almost no prevalence of deficiency in the west, with studies showing prevalence at below 1% (Bird, 2017; Hanson, 2016). The problem lies in the widespread practice of grouping fat-soluble vitamins together without adequate harm reduction education, effectively treating vitamins with vastly different safety profiles as equivalent (Youness, 2022).
Unless there is a clear, documented deficiency measured, supplementation is not beneficial and in most cases simply detrimental.
Omega-3 — 2000 mg
The polyunsaturated fatty acids, omega-3, particularly EPA and DHA, serve distinct neurological and cardiovascular functions. EPA primarily drives anti-inflammatory pathways and mood regulation, while DHA comprises roughly 40% of polyunsaturated fatty acids in the brain and is essential for membrane fluidity and neurotransmission (Cholewski, 2018).
The omega-3 index provides a useful biomarker for sufficiency. It’s calculated by measured EPA and DHA as a percentage of red blood cell fatty acids. Most Western populations show indices below the optimal 8%, with typical values ranging from 4-6% (Schuchardt, 2022). This widespread insufficiency reflects both low fish consumption and the dominance of omega-6 fatty acids in modern diets (Dempsey, 2023).
In ADHD populations, meta-analyses show modest but consistent benefits from supplementation, particularly for hyperactivity and attention measures (Bloch & Qawasmi, 2011; Chang, 2018). The effects appear most pronounced in those with lower baseline omega-3 status, suggesting a threshold effect rather than linear dose-response (Hawkey & Nigg, 2014).
For depression, the evidence is strongly differentiated between EPA and DHA. EPA shows small-to-moderate effects as monotherapy, with larger benefits when used adjunctively with conventional antidepressants (Liao, 2019; Appleton, 2021). Meta-analyses consistently find EPA-dominant formulations (>60% EPA) outperform balanced or DHA-heavy preparations for mood outcomes (Carnegie, 2024). The mechanism likely involves reduction in neuroinflammation and enhanced monoamine transmission (Larrieu & Layé, 2018).
Bioavailability varies significantly by form. Natural triglyceride forms found in fish oil demonstrate superior absorption compared to the ethyl esters commonly used in pharmaceutical preparations (Chevalier, 2021; Chevalier & Plourde, 2021). Re-esterified triglycerides fall between these extremes (Maki & Dicklin, 2019; Minton, 2023). Since most clinical trials use the less bioavailable ethyl ester forms, real-world benefits from quality supplements may exceed published effect sizes.
Both EPA and DHA have half-lives of 2–3 days, with EPA being slightly longer on average (Zuijdgeest-van Leeuwen, 1999; Braeckman, 2014). Differences in absorption between triglyceride and ethyl ester forms were less pronounced after subjects ate high-fat meal compared to a low-fat meal, suggesting that absorption is increased alongside dietary fat (Schuchardt, 2011).
Safety concerns emerge only at very high doses (>3g EPA), where antiplatelet effects may increase bleeding risk (Javaid, 2024). For most individuals, 2000 mg total omega-3 with an EPA:DHA ratio favoring EPA, at a minimum of 60:40, represents a reasonable balance between efficacy and safety. Red blood cell omega-3 index testing can confirm adequate supplementation only after 8-12 weeks, as cellular turnover creates this lag between intake changes and measurable outcomes (Schuchardt, 2022).
Bupropion ER
In the treatment of depression, commonly adjunct to psychotherapuetic interventions are antidepressants. However most antidepressants often barely outperform placebos (Huedo-Medina, 2012; Kirsch, 2008). Notably, bupropion is one of the few exceptions (Patel, 2016).
Its mechanism as a NDRI reduces the rate at which NE/DA is reabsorbed into presynaptic neurons, increasing the availability for signalling within neurons (Stahl, 2004).
Unlike typical first line antidepressants SSRIs or SNRIs, it has minimal serotonergic activity. This means that common but poorly tolerated side effects like weight gain and sexual dysfunction are avoided (Jain, 2002; Anderson, 2002; Mouawad, 2025; Higgins, 2010).
Though it is a weak inhibitor compared to other drugs it remains useful off-label in second-line or adjunct treatment of ADHD (Reimherr, 2005; Wilens, 2001), with a compounded benefit when comorbid (Daviss, 2001).
An additional aspect of bupropion is antagonism of the nicotinic acetylcholine receptors, which is responsible for its use in smoking cessation treatment (Hurt, 1997). This is particularly important within the context of depression and ADHD as both are risk factors for substance use disorders (SUD), especially nicotine (Glassman, 1990).
The proprietary extended release formulation, Wellbutrin XL, is the only antidepressant currently formally approved for SAD prophylaxis in the U.S (FDA, 2022). Its approval was supported by RCTs showing reduced recurrence of depressive episodes when started in autumn (Modell, 2005), with later meta-analysis confirming preventive efficacy in high-risk patients (Niemegeers, 2013).
Lisdexamfetamine
The most effective ADHD medication, lisdexamfetamine (LDX), is a prodrug which means it gets converted in vivo to its active form.
Specfically, it is (dextro)amphetamine attached to a lysine amino acid. The conversion is a pretty simple enzymatic reaction where the lysine-amphetamine peptide bond gets hydrolysed by red blood cell peptidases (Pennick, 2010). The rate of the process is based on enzyme availability rather than route of administration.
This means that even if you were to crush and insufflate, or even take it intravenously, it wouldn’t result in a stronger ‘high’. This is why regulators tend to like it, because it needs to be converted organically (supposedly) preventing abuse (Krishnan & Stark, 2008). Ironically, the only real requirement for extraction isn’t a biochemistry degree, but an internet connection and two hours of hyperfocus.
When compared to taking two immediate release (IR) doses in succession; the real benefit of LDX becomes clear. It has a flatter dose-response curve (Strajhar, 2019), meaning slower consistent release and steadier plasma levels and smoother effects (Ermer, 2010; Strajhar, 2019).
Though it’s double edged, metabolism varies significantly between individuals, some ADHD patients need a faster onset or shorter duration (Childress, 2019). Some countries (UK, Australia) routinely prescribe dextroamphetamine IR for dose titration and breakthrough coverage, while others pretend it’s somehow more dangerous than the prodrug version of the exact same molecule.
In the U.S., a different form of amphetamine is used more frequently — Adderall, or mixed amphetamine salts, which contains 25% dextroamphetamine sulfate, 25% dextroamphetamine saccharate, 25% racemic amphetamine aspartate, and 25% racemic amphetamine sulfate (Heal, 2013). Don’t ask me why they chose to put four different forms inside, but overall this is a 3:1 ratio of dextroamphetamine to levoamphetamine.
Dextroamphetamine and levoamphetamine are both the same molecule but mirror images which can’t be superimposed. When light hits either molecule they polarise light differently, this is called chirality. Each enantiomer (or optical isomer) of the chiral pair has a prefix denoting which type: levo- prefix sometimes abbreviated as
Stimulants will always be linked to misuse; but in the case of ADHD they’re so effective that they remain the primary The irony is self-evident: substances prescribed for use in patients with comorbid substance use disorders. The effect is recursive: reducing impulsivity and need for self-medication in turn lowers the very risk factors making prescribing stimulants seem risky (Konstenius, 2010; Levin, 2015).
Effectiveness is not uniform across different pyschostimulants (Cortese, 2018). In countries with stricter laws surrounding the use of stimulants, even as medication, methylphenidate (MPH) is predominantly used. MPH shows efficacy inbetween amphetamine’s enantiomers (Smith & Davis, 1977), but is worse tolerated. Atomoxetine (ATX) an sNRI showed a lower effectiveness relative to LDX at 65% versus MPH ER’s 81% (Roskell, 2014). Modafinil is commonly grouped alongside whilst being a wakefulness promoting agent rather than a psychostimulant, meaning it has less effect on the central nervous system; it typically shows no improvement against placebo (Arnold, 2014).
There is often a mild-moderate depression or mood improvement ADHD symptom reduction (Adler, 2009). SSRIs can sometimes worsen apathy or bluntness in ADHD because serotonergic agonism (notably of the 5-HT2A/2C receptors) can suppress dopamine release (Millan, 1998; Porras, 2002).
The enzymatic machinery that converts lysine-amphetamine to its active form is only the beginning (Roncero & Alvarez, 2014). Sustained catecholamine release creates demands that may outpace the body’s ability to replenish what’s consumed — though documenting this clinically remains frustratingly elusive.
Vitamin B12 (Methylcobalamin) — 500 µg
The methylation cycle is an integral part of intracellular biology. B12 is a cofactor for two key reactions: the conversion of homocysteine to methionine (via MTR) and norepinephrine to epinephrine (via PNMT).
Higher income countries generally do not have severe B12 deficiencies, though marginal deficiency is displayed at a reasonable enough rate to be significant. In data from the U.S. department of agriculture as much as ~20% of the healthy individuals sampled were below the cutoff used for identifying replete status (Allen, 2009). Similar rates were found in a cross sectional study for prevalence of deficiency amongst corporate employees (Chirravuri, 2023).
Multiple pediatric ADHD studies show significantly elevated homocysteine and reduced B12 levels, indicating functional disturbance in the methylation cycle (Altun, 2018; Lukovac, 2024). Meta-analysis showed this was also prevalent in adults (Tan, 2023). Similar links were found for depression (Esnafoglu & Ozturan, 2020).
There is empirical evidence that high DA/NE turnover from stimulants increases vitamin B12 demand (Madebo, 2022).
Unlike folate, B12 has no established upper limit and overdose risk is negligible, making it safer to hedge high. For most individuals, high-dose oral is effective; injections are reserved for clinical deficiency or absorption disorders (Abdelwahab, 2024).
Particularly synthetic cyanocobalamin has poor conversion and safety; therefore hydroxo-, methyl-, or adenosylcobalamin should be preferred (Paul & Brady, 2017). This, however, this is minor consideration compared to the real concern: ensuring B12 sufficiency, since hypovitaminosis manifests subtly but carries major neurological consequences. Practically 500µg orally is cheap, widely available, and well above the RDA.
Vitamin B9 (L -methylfolate) — 400 µg
Similarly central in methylation, folate or B9 is also crucial in synthesis of DNA and the production of red blood cells (Bailey, 2015).
Prevalence of inadequate folate levels in Germany was found to be roughly 14% (Mensink, 2016), additionally some studies have shown that incidence of ADHD is inversely correlated (Landaas, 2016).
Multiple studies show that B9 as monotherapy or an adjunct to existing antidepressant therapy shows a statistically signficant improvement in depression symptoms (Shelton, 2013; Esnafoglu & Ozturan, 2020; Hardin & Baldwin-Sayre, 2020; Carboni, 2022; Maletic, 2023).
There is considerable talk about MTHFR genetic mutations from genetic testing companies. Whilst they suggest that testing is essential, evidence is low that B9 intake is disrupted enough to have large clinical implications (Maruf, 2021; Carboni, 2022). Regardless, it’s not harmful nor expensive to avoid synthetic folic acid in favour of natural forms like
B12 sufficiency is measured by plasma homocysteine (Carmel, 2008). Whilst excess folate intake isn’t directly harmful, it can normalise hematologic markers potentially masking B12 deficiencies (Tucker, 1996). To minimise risk, it is important to avoid over-supplementation and stay near the RDA ~400 µg.
Vitamin B6 (P5P, Pyridoxal 5’-Phosphate) — 20 mg
Dopamine and serotonin biosynthesis pathways are gatekept by PLP availability: the most important being aromatic
Deficiency is uncommon in wealthier countries (citation needed). Some research may indicate that particular risk factors for PLP deficiency include a weak association with depression (Esnafoglu & Ozturan, 2020) and a strong correlation with prevalence of ADHD (Landaas, 2016; Altun, 2018; Hunter, 2025).
There is a considerable debate online surrounding B6 toxicity (Apostolopoulos & Feehan, 2025). It’s overstated, but also easy to see where the panic comes from. Regulators don’t even agree on the daily maximum intake. The EU pushed a much lower upper-limit than the US, 12 mg and 100 mg respectively which is hardly an insignificant difference (Institute of Medicine, 1998; EFSA, 2008; EFSA, 2023; Schellack, 2025).
National pharmacovigilance databases keep getting PN heavy reports, even at modest doses, because most consumer focused B-complex products are Pyridoxine Hydrochloride by default; Vrolijk (2020) with data from the Netherlands Pharmacovigilance Centre Lareb found that <3.5% of subjects who reported toxicity were taking PLP. This reinforces the suggestion that not enough consideration is being taken with regards to the varying tolerability of the different vitamers.
Mechanistically, accumulation of PN competes with PLP, which can inhibit pyridoxal kinase, reducing GABA synthesis, leading to sensory neuropathy such as nerve degeneration (Hadtstein & Vrolijk, 2021). Though it should not be any real cause for concern.
If hypovitaminosis is known, or demand is plausibly higher, 20 mg P5P is a defensible cofactor-support dose. It aligns with the aforementioned PN-vs-PLP safety asymmetry but remains a moderate dose above bare-minimum intake, though far below pharmacologic territory.
Magnesium L -Threonate — 1000 mg
In the brain, neurons usually idle at -70 mV; they’re affected by two main systems: GABA hyperpolarises, reducing the potential, and NMDA/glutamate the inverse, depolarising and increasing the potential. When the voltage passes the -55 mV threshhold, the the neuron fires and is activated (Khadria, 2022). Many sedatives such as benzodiazepines, e.g. alprazolam/Xanax, work by increasing GABA (Swinson, 1987; Ballenger, 1990; Stachowicz, 2008; Perkins, 2013), whilst stimulants increase glutamate via DA/NE (Reid, 1997).
Magnesium sits in the entrance to the NMDA receptors and blocks ions from depolarising the neuron (Coan & Collingridge, 1985). This is important because the at rest balance of these system determines the sensitivity. Neuron hyperexcitability is tightly coupled to presentations of effects such as depression, anxiety/panic disorders (Niswender & Conn, 2010), and muscle tension.
Whilst commonly disregarded by the physicians prescribing ADHD medication, muscle tension, especially bruxism (tension in the jaw) is another noticable but irritating side effect (Pavlou, 2023). Elevated epinepherine/dopamine whether from exercise, stress, or stimulants—shift magnesium intracellularly, lowering plasma levels; this effect is well documented in exercise physiology (Joborn, 1985), and even but virtually unexamined in ADHD medication use, despite the shared mechanism (Fratto & Manzon, 2014).
Systematic review suggested that there may be some correlation between ADHD and magnesium levels though results were mixed (Botturi, 2020). Both a significantly higher prevalence of magnesium deficiency and a significant decline in depression symptoms with supplementation has been reported (Pickering, 2020; Moabedi, 2023).
The blood-brain barrier or BBB regulates the transfer of other chemicals from blood into the CNS where it’s carried by the CSF (cerebrospinal fluid). The level of magnesium in the CSF stay particularly stable, only slowly changing over time (Ghabriel & Vink, 2011). This becomes of particular importance when you consider the salt forms of which magnesium can be taken which vary greatly in both absorption and BBB permeability (Romeo, 2019; Mathew & Panonnummal, 2023; Cazzaniga, 2022; Koc, 2025). This means that forms with lower immediate bioavailility to the brain are less beneficial for the acute effects but remain useful for consistent usage.
L-Threonate shows promise for high BBB penetration in new research. Most of the human studies are funded by the company who owns the patent, directly or indirectly by those who produce and/or sell it (Hausenblas, 2024; Zhang, 2022). Outside of industry funding, human studies remain limited, but there is a reasonable amount of evidence to support it in animal models (Fu, 2025; Shen, 2019; Xiong, 2022).
Hypermagnesemia rare unless without severe renal impairment (less than 1/3 of function), as otherwise excess is cleared quickly by the kidneys (Kim, 1969; Elin, 1988). It’s important to note the difference between elemental magnesium and total salt, usually the actual magnesium content sits at roughly 20% of the total mass though this varies by form (Firoz & Graber, 2001; Ranade & Somberg, 2001; Uysal, 2019).
The main risk of harm is gastrointestinal discomfort at high doses (Guerrera, 2009) but this can be alleviated by splitting doses throughout the day. Taking half of this as 1000 mg Magnesium L-Threonate is reasonable. If you do take ADHD medication, timing it roughly one hour after ADHD medication allows synergy making use of the more noticable acute effects.
L -Tyrosine — as needed
As the rate-limiting precursor in the catecholamine biosynthesis,
Studies show cognitive benefits in certain circumstances, particularly when sleep deprived, under high stress or cognitive load (Deijen, 1999), and with stimulant use (Eisenberg, 1988; Woods & Meyer, 1991). Additionally there is evidence it may help with convergent (deep) thinking, though has no impact on divergent (creative) thinking (Colzato, 2015). As a note, the medical establishment tends to use the latin ‘PRN’ which just means ‘as needed’.
When the system is stable there is no need for extra precursors as the enzymes produce normally. Enzymes typically rate limit themselves (likely via feedback inhibition), therefore excess supplementation of
Absorption typically competes with other amino acids (Fernstrom & Wurtman, 1971), so taking it away from proteins is preferred. It may be potentially stimulating so in the morning is preferred. It’s better for targeted use as its half-life is 1–2 hours. Clearance is roughly 5–6 hours. Dosing around 500–1000 mg L-Tyrosine is common (Rucklidge, 2009).
Intermission
We’re halfway through, you could just quit here though; the morning group covers the most crucial supplements.
‘psst, hey kid, want some chelated minerals? I got the good stuff — bisglycinate, real bioavailable shit’
— me probably
Anyway, more to follow, timing matters because pharmacokinetics exists.
Evening
Magnesium Taurate — 1000 mg
Along with its CNS-related effects, magnesium is also important for cardiovascular health. The detrimental effects of hypomagnesia are well documented, showing sigificantly potential for poorer clinical outcomes (Fritzen, 2023).
Chronic and a subclinical magnesium defiencies are becoming more common due to widespread dietary changes and a reduction in magnesium in food crop; in the U.S. roughly 43% of people were below the RDA (Beltsville Human Nutrition Research Center, 2020). Though slightly lower, research estimated prevalence in Germany at 26–29% (Micke, 2020).
Deficiency is widespread and has significant health implications. There is strong evidence that hyomagnesia may be associated cardiovascular diseases (Gums, 2004; Tangvoraphonkchai & Davenport, 2018; Nielsen, 2024). It can trigger arythmias and increase blood pressure, straining the heart and increasing the potential for heart attack or stroke (DiNicolantonio, 2018). Therefore it is even more critical to prevent deficiency in those using ADHD medication which already strains the heart.
As noted previously, there are many salts of Mg: (bis)glycinate,
Following from the split dose of magnesium, it’s sensible to match the elemental content at 1000 mg of magnesium taurate which primarily cardiovascular outcomes versus
Individual tolerance varies and some may benefit from higher doses beyond 400 mg elemental magnesium total as long as GI tolerance allows it.
Vitamin C (Ascorbic acid) — 500 mg
Ascorbic acid reduces oxidative stress and supports white blood cell functions specifically neutrophils, among other things (Bozonet & Carr, 2019). In the west, supplementation is typically not needed, however it can increase absorption of both Zinc and Calcium (Sandström & Cederblad, 1987). Additionally, mega-dosing (2g) at the onset of and during the common cold has shown to reduce symptom duration by up to ~8-12% (Hemilä, 1999). There is essentially no risk of toxicity and it’s well tolerated even at high doses.
Zinc Bisglycinate — 50 mg
Studies has been shown that the highest levels of zinc intake reduced the risk of depression by up to ~28% (Yosaee, 2022).
Cochrane’s review by Nault (2024) concludes that zinc has ‘little to no effect’ relies on absolute differences in cold duration. This, however, is a metric that doesn’t fit the data well — Hemilä & Chalker (2024) point out that using a relative scale, which better reflects variability in illness duration, consistently shows a meaningful reduction in duration (~10–15%) with supplementation above 200 mg/day.
This works synergistically with vitamin C, whereby the combination was shown to reduce common cold symptom duration further up to ~9–27% (Wintergerst, 2006; Maggini, 2012). In some studies zinc’s role brought this up to even ~50%; of note is the form remains important for acute effects, it must be dissolved in the mouth as a lozenge, typically zinc monomethionine or acetate (Prasad, 2008). Hereby it is also superdosed taking a 10-20 mg lozenge every 2–3 hours, normal supplementation should be paused for this duration to lower the risk of adverse effects from having excess zinc.
Zinc Bisglycinate for daily supplementation to maintain zinc status is preferred as it is highly bioavailable and well tolerated (Devarshi, 2024). Typically 25-50 mg elemental zinc is used, this should be balanced with copper intake but 25 mg is a reasonable bet. Also the amount listed for zinc is generally elemental, don’t ask me why this is inconsistent.
5-HTP (5-Hydroxytryptophan) — as needed
The aromatic L-amino acid decarboxylase (AADC) enzyme synthesises serotonin (SE), or 5-HT, primarily from 5-HTP. There are many similarities to
Supplementation carries a risk of serotonin syndrome, which is a potentially fatal condition whereby primarily the 5-HT2A subtype of the serotonin receptors are overstimulated causing a number of harmful symptoms; this risk is serious and shouldn’t be overlooked, but if you understand the causes you can safely use supplements (Boyer & Shannon, 2005).
-
5-HTP should not be mixed with the following: antidepressants which prevent serotonin reuptake into the synapse such as SSRIs and SNRIs (Novella, 2025), and monoamine oxidase inhibitors (MAOIs) which typically inhibit the MAO-A enzyme responsible for the breakdown of serotonin (Chamberlain & Baldwin, 2021; Chamberlain & Baldwin, 2022).
While bupropion has minimal direct serotonin activity, regulatory agencies have added warnings about serotonin syndrome risk when combined with other serotonergic agents. Exercise caution if combining 5-HTP with bupropion.
-
Discontinuing these medications isn’t enough and there is a significant cool-down window for the system to stabilise (Keks, 2016).
A recent review from Badar (2024) shows further interactions in more detail.
-
Significant care should be taken with dosing as higher doses of 5-HTP is enough to induce serotonin syndrome alone (Turner, 2006).
There are many considerations in order to avoid high amounts of systemic serotonin — despite this, 5-HTP remains a useful tool.
Most meta-analyses underestimate 5-HTP’s effect because they pool trials with low dosing, short durations, or mixed populations without serotonin deficiency (Byerley, 1987). Layered studies such as (Pöldinger, 1991) show outcomes closer to SSRIs in responsive patients.
Dopamine and norepinepherine drive task engagement and reward learning, but behavioral inhibition and impulsive aggression are also modulated by serotonin in frontostriatal and orbitofrontal circuits (Oades, 2008). Experimental serotonin depletion reliably increases impulsive choices and disinhibition in some paradigms (reversal learning, response inhibition, delay discounting), supporting a link between low serotonin levels and impulsivity (Walderhaug, 2007; Dalley, 2011).
ADHD samples are heterogeneous, but reviews note serotonergic abnormalities in subsets, particularly those with impulsive subtype profiles rather than inattentive-only (Skirrow & Asherson, 2013). Within this frame, 5-HTP may be used as a mechanistic tool, bypassing tryptophan hydroxylase, to raise 5-HT synthesis when baseline tone is low.
There no RCTs directly targeting ADHD with 5-HTP, but pathway is biologically plausible given current models of impulsivity (Bari & Robbins, 2013). Mismatch between DA:SE ratios may, however, worsen apathy or cognitive flexibility in some, suggesting this would not be universal (Robbins, 2012).
As serotonin itself is the precursor to melatonin, 5-HTP has a potential to increase tiredness, and should be dosed only after midday for sleep benefits (Silber & Schmitt, 2010).
Absorption is slow at roughly 2–3 hours orally, with a half-life of 5 hours (Magnussen & Nielsen-Kudsk, 1980). Common dosage range in trials is 50–300 mg daily, sometimes divided.
Melatonin — as needed
The hormone, melatonin, is most widely used for its important chronological function, namely: adjusting sleep patterns, whether generally or for insomnia. Alongside its primary purpose it possesses strong neuroprotective (Won, 2021) and anti-oxidative properties (Pandi-Perumal, 2006).
The rate of melatonin production is controlled by the SCN, which functions as the body’s master clock (Reiter, 1993). Light entering the retina synchronizes the SCN, which regulates norepinephrine release from sympathetic fibers onto the pineal gland, triggering melatonin synthesis (Simonneaux & Ribelayga, 2003). Bright light therapy indirectly modulates systemic melatonin using environmental luminance via this mechanism.
Systemic review shows that melatonin has a signficant improvement in sleep onset latency (SOL), duration, but not in frequency of wake events. Patients with comorbid ADHD experienced smaller improvements in SOL (approximately 13% less reduction) compared to those with primary sleep disorders (Salanitro, 2022); this discrepancy may be linked to catecholaminergic dysregulation, as is common in ADHD (Palm, 2021).
Efficacy of medications which affect NE or in any capacity as an NRI may be measured using aMT6 metabolite of melatonin as a biomarker (Tonon, 2021).
In seasonal depression, the majority of patients have a substantial delay in circadian phase, and correction via both supplementation and light therapy show improvement in these patients; regardless intensity of SAD had a parabolic association with phase difference with both early and late sleep times showing stronger symptoms (Lewy, 2006).
Interestingly, another one of the few antidepressants which shows significant difference vs placebo is the melatonin agonist, agomelatine (De Bodinat, 2010). Since it also possesses 5-HT2C antagonist properties (disinhibiting DA/NE release) it may appear almost designed specifically to treat SAD, where the depression is specifically linked to circadian rhythm, addressing both sides mechanistically (Kasper & Hamon, 2009); however, light therapy still dominates in treatment and more research is needed specifically indicate agomelatine for SAD. Evidence on long-term efficacy remains limited; a systematic review found small effect sizes and high heterogeneity across studies (Koesters, 2013).
Melatonin has been indicated as a safer alternative of treating sleep disorders using sedatives and/or anxiolytics, like benzodiazepines or Z-drugs, as it has no known risk of dependence, significant tolerance, or serious adverse affects (Cardinali, 2012).
There is no established toxic dose in humans; though cochrane’s review from Herxheimer & Petrie (2002) suggested that 5 mg melatonin is measurably more effective for fixing jet lag than 0.5 mg. Meta-anaylses of studies using higher than 10 mg note no increase in serious side effects, and only a 40% increase of minor side effects such as drowsiness, headache, and dizziness (Menczel Schrire, 2022). Studies exist with very high doses (100 mg IV, 500–1000 mg oral) indicating a reasonable level of safety overall (Andersen, 2016; Valcavi, 1987).
Dosing appears to be particularly nuanced, there is high variance between individual bioavailability, ranging from 10–56% orally due to first-pass metabolism (Harpsøe, 2015). Sublingual and buccal forms bypass this, leading to a faster onset and more predictable level in plasma; while sustained-release tablets maintain low, consistent levels throughout the night — mimicking physiological patterns (Zisapel, 2018).
Timing of use is considerably more important; there are three ways to achieve different effects. Melatonin can act as a sedative by dosing 30–60 minutes before bedtime. Taking it in early evening, roughly 6–8 hours before sleep time, advances the circadian rhythm. Delay is achieved by taking melatonin shortly after waking in the morning (Lewy, 1992).
For restoring circadian rhythm in the case of jet lag or mild-to-moderate insomnia, 5–10 mg appears to be a good bet, 0.5–5 mg are sufficient for phase shifting. Individual differences in metabolism via CYP1A2 (fast versus. slow metabolisers) may require higher doses for optimal effectiveness. Many commercial formulations contain pyridoxine (synthetic B6), make sure to avoid these.
L -Theanine — as needed
An amino acid found in tea leaves,
It appears most effective for acute use, e.g., situational anxiety or for sleep; chronic daily dosing shows diminishing returns likely due to receptor adaptation. A typical dose is 200–400 mg. It’s often combined with caffeine to reduce stimulant related jitteriness (Williams, 2020). The half-life is short at roughly 1–2 hours, so timing is important.
What’s Not Here (And Why)
If something’s not included, it’s because:
- Insufficient evidence (NAC, nootropics, lion’s mane)
- Requires testing first (iron, vitamin A)
- Redundant (you don’t need five forms of magnesium, two is already pushing it)
- Too individually variable (ashwagandha, glutamate modulators)
- Expensive placebo (anything with lightning on the bottle)
- Prescription drug in disguise (lithium, kratom)
- Don’t like the name (st. john’s wort)
Integration
The overall system can be described succinctly as follows:
The metabolic/cofactor system: B12/B9/B6 for methylation, creatine for energy buffering, creating a fault-tolerant configuration (high B12, low B9, P5P not PN) which works regardless of MTHFR status or metaboliser type.
The catecholamine support system: tyrosine → magnesium → zinc chain, where stimulants deplete downstream cofactors. Morning magnesium handles acute muscle tension, evening dose supports cardiovascular stress, and C/zinc back up immune function that NE disrupts.
The circadian/mood system: D3/K2 for a seasonal baseline, melatonin for phase correction, 5-HTP for serotonergic balance, creatine and omega-3 for inflammation.
Testing
Becoming obsessive with bloodpanels is unlikely to significantly improve your outcomes, but is certain to cost a considerable amount of money. As previously mentioned, the main idea of this system is the ability to do it blind.
In most cases they should not be necessary but some tests may still be useful: Check ferritin for an iron deficiency (correct with supplementation), copper and zinc for an inequal ratio (correct by adjusting zinc dosage), and B12, if you are vegetarian or vegan.
Afterword
To adherence easier, filling a pill organiser weekly reduces resistance. This stack costs roughly €50/month, takes 30 seconds each morning/evening, and handles most of the biochemical friction that makes ADHD and modern life harder than it needs to be.
Ok good? Now stop reading about supplements and go do something interesting.