Blog

Building Your First Supplement Stack

Unfair Team • January 11, 2026

Build stacks like experiments: pick one goal, change one variable at a time, and keep only what produces a meaningful, repeatable benefit at a risk level you accept.

The mistake most people make is not “choosing the wrong supplements.” It is building a stack that is impossible to interpret. If you cannot tell what caused the change, you cannot improve it. Your first stack is about signal.

When not to stack at all

Do not start a stack when the right next step is diagnosis, not optimization. Stacks are useful when you have a stable baseline and you want to test a specific lever. They are a poor substitute for figuring out why something is happening.

Pause stacking and address the underlying issue first if any of these are true:

• New or rapidly worsening symptoms (sleep, mood, pain, GI changes, fatigue) over weeks.
• Severe sleepiness, loud snoring, or observed apneas. Supplements will not fix sleep apnea.
• Panic attacks, suicidal thoughts, or severe depression. Treat this as urgent, not “stackable.”
• Unexplained weight loss, blood in stool, persistent vomiting, black stools, or fevers.
• Chest pain, fainting, palpitations, new severe headaches, or neurologic symptoms.
• Pregnancy or trying to conceive, unless the plan is explicitly pregnancy-safe.
• You are starting or changing prescription meds this month. Let one change settle before adding noise.

Also do not stack if you are unwilling to measure anything. If you will not track, a “stack” becomes a collection hobby.

The one-goal, one-stack framework

A stack is a short protocol with a single purpose. You build it in layers, but you only change one layer at a time.

The stack layers

• Foundation (optional) Only corrects a likely deficiency or structural gap that would otherwise distort your results (for example: vegan B12, low vitamin D, very low omega-3 intake). This is not “take everything.”
• Goal-specific (required) One supplement with the best evidence-to-risk ratio for your goal.
• Support (optional) Added only if it improves tolerability or adherence without changing the goal (for example: split dosing to reduce side effects). If it has its own effects on the goal, it is another variable and needs its own test.

The decision rules

You need explicit rules before you begin. Otherwise you will keep adding things until you feel “something,” and then you will not know what worked.

A simple cycle that works

1. Baseline Track your endpoint for 7–14 days with no supplement changes.
2. Intervention Add one supplement at a conservative dose and fixed timing.
3. Review Compare the last 7 days of baseline to the last 7 days of the intervention period.
4. Decision Keep, adjust dose, or remove based on the rules below.

Personalize by goal category

Use this section to choose what to test first. The goal is not to find “the best supplement.” It is to choose the best first experiment.

A practical ranking for first-stack candidates is:

1. Strong evidence + low interaction risk + predictable timing
2. Moderate evidence + manageable risk
3. Weak or population-specific evidence (only after you have measurement discipline)

Sleep quality

Start by deciding what is actually broken:

• Sleep onset (you cannot fall asleep)
• Sleep maintenance (you wake up and cannot return)
• Circadian timing (your sleep is shifted late or early)

First-stack candidates (pick one):

• Melatonin for sleep timing and sleep onset, especially when circadian timing is the problem. Effects on primary insomnia are modest in meta-analyses. Use it as a timing tool, not a sedative. ^{1 2}
• Magnesium can help some people, especially when intake is low or in older adults with insomnia in small trials, but the overall interventional evidence is not uniformly strong. ^{3 4}

Avoid stacking multiple sedating agents on night one. If you do, you will not know what caused grogginess or vivid dreams.

Anxiety and stress

Separate state stress (acute) from baseline anxiety (persistent). Many “calming” stacks blur these.

First-stack candidates:

• Ashwagandha has randomized trials showing improvements in stress and anxiety measures in some populations over ~6–8 weeks, but product standardization and safety considerations matter. ⁵
• L-theanine (acute) has evidence for small acute effects on calmness and attention-related outcomes, especially when paired with caffeine in studies. ⁶

If you take antidepressants or have bipolar-spectrum symptoms, be conservative with mood-active supplements and prioritize measurement and stop rules.

Focus and ADHD-like symptoms

First decide whether the issue is sleep debt, under-fueling, too much caffeine, or task design. If you fix those, “focus supplements” often become unnecessary.

First-stack candidates:

• Caffeine has consistent evidence for improved alertness and exercise performance. For cognition, it is reliable but tolerance and sleep disruption are common failure modes. Use dose discipline. ^{7 8}
• Caffeine + L-theanine has randomized and meta-analytic evidence for improved aspects of attention and alertness compared to placebo, with a “calmer” profile in some studies. ^{6 9}

Do not stack multiple stimulants. If caffeine is already high, the first “stack change” is often reducing or timing it.

Mood

Treat mood like a slow variable. Most supplements that plausibly affect mood do not act in 48 hours.

First-stack candidates:

• Omega-3 (EPA/DHA): evidence is mixed by population and formulation, and effects are not universal. If you test it, run a longer trial and track with a validated tool like PHQ-9. Omega-3 also has a clearer role for triglycerides than for mood. ¹⁰
• Vitamin D: if you are deficient or at risk, correcting status is reasonable, but supplementation in vitamin D-replete individuals is not consistently linked to mood benefits. Treat it as “correct a gap,” not a nootropic. ¹¹

If mood symptoms are severe, do not treat this as a supplement problem.

Strength and hypertrophy

This category is unusually forgiving because the endpoints are easier to measure and the evidence base is stronger.

First-stack candidates:

• Creatine monohydrate: one of the best-supported performance supplements for high-intensity work and lean mass gains during training. Timing matters less than daily consistency. ¹²
• Protein supplementation if total protein intake is low. Meta-analysis suggests diminishing returns beyond roughly ~1.6 g/kg/day when supplementing during resistance training in healthy adults. ¹³ (Also see ODS guidance for athletes.) ⁷

If you do not train progressively, no stack will rescue this goal.

Endurance

Pick the endurance domain:

• Long steady-state
• Intervals and repeated efforts
• Race-day performance vs training volume

First-stack candidates:

• Caffeine: consistent ergogenic effects in many contexts, commonly used 15–60 minutes pre-exercise, with meaningful inter-individual variability. ^{7 8}
• Dietary nitrate from beetroot juice: studied protocols often use a single dose ~2.5–3 hours pre-exercise or short daily loading phases. Effects vary by individual and context. ⁷
• Beta-alanine: most relevant for efforts roughly 1–4 minutes and repeated high-intensity work. Requires weeks, not days. ¹⁴

Recovery and soreness

Recovery stacks fail when people track the wrong outcome. The goal is usually not “less soreness.” It is “return of performance” and “ability to train again.”

First-stack candidates:

• Adequate protein to support recovery, especially if baseline intake is low. ^{13 7}
• Tart cherry has meta-analytic evidence for small improvements in soreness and recovery markers, often tested around intense blocks. ^{15 7}

Joint pain and inflammation

Joint pain is not one thing. A stack that helps osteoarthritis discomfort may not help inflammatory arthritis.

First-stack candidates:

• Curcuminoids: meta-analyses in knee osteoarthritis suggest pain relief with both low- and high-dose approaches, but formulations vary and safety signals exist for high-bioavailability products. ^{16 17}
• Omega-3 (EPA/DHA): may be reasonable to test when inflammation is part of the picture, with attention to bleeding risk and medication interactions. ¹⁰

Glucosamine and chondroitin may help some people with osteoarthritis, but results vary and warfarin interaction risk matters. ¹⁸

Gut comfort

Gut stacks fail when people treat “gut comfort” as a single endpoint. Decide whether you are targeting:

• Stool form and frequency
• Abdominal pain and cramping
• Bloating and gas
• Reflux and nausea

First-stack candidates:

• Soluble fiber (psyllium) for global IBS symptoms and stool normalization, with evidence for lipid improvements at typical doses. Increase slowly. ^{19 20}
• Enteric-coated peppermint oil for IBS symptoms in the short term, with guideline and meta-analytic support. ^{19 21}

Metabolic health

This category should be mostly measurement. Without labs or repeated readings, you cannot tell if anything changed.

First-stack candidates:

• Psyllium: evidence supports meaningful LDL reductions at typical doses in trials and can help with satiety and post-meal glucose dynamics. ²⁰
• Omega-3: clearer role for lowering triglycerides at higher EPA/DHA intakes, including prescription-level dosing in clinical contexts. ¹⁰

Berberine has meta-analytic evidence for glycemic improvements in type 2 diabetes populations, but it can meaningfully interact with diabetes medications. Treat it as “drug-like” and do not self-stack it casually. ^{22 23}

General micronutrient sufficiency

Do not guess. “Micronutrient sufficiency” is best handled with diet assessment plus targeted labs where appropriate.

Common high-yield cases:

• Vitamin D limited sun exposure, darker skin at high latitudes, or known low 25(OH)D. ¹¹
• Vitamin B12 vegan or near-vegan diets, long-term metformin use, or low B12 status. ²⁴
• Iron heavy menstrual bleeding, endurance athletes, restrictive diets, or confirmed low ferritin. Supplementation should be lab-guided. ²⁵

A broad multivitamin can be reasonable when diet quality is poor, but it is not a substitute for fixing the diet pattern driving the gaps.

Measurement: what to track and how long to run it

Pick one primary endpoint per goal. Add one secondary endpoint only if it helps interpret side effects or tradeoffs.

Practical tracking defaults by goal

Control the confounders that matter

You do not need perfect control. You do need stability.

• Keep sleep schedule within a 60–90 minute window.
• Keep caffeine timing stable and avoid adding new stimulants.
• Keep training structure stable for performance goals.
• Keep alcohol stable because it can swamp sleep and recovery signals.
• Keep diet pattern stable if the supplement works through digestion or energy.

If you cannot keep these stable, extend the trial length rather than adding more ingredients.

Dosing and timing: how to choose a dose and schedule that actually tests something

A good first-stack dose is not the highest dose. It is the lowest dose that is plausible to work based on human trials and is tolerable enough to stay consistent.

Choose a dose range the right way

1. Start with studied ranges, not influencer doses.
2. Pick the low end if side effects are common.
3. Only titrate upward if:

• adherence is high, and
• side effects are minimal, and
• you have not reached your pre-set success threshold.

Examples of evidence-grounded ranges and timing:

• Creatine monohydrate daily 3–5 g is a common maintenance approach, with optional loading protocols (higher short-term dosing) to saturate faster. Consistency matters more than timing. ¹²
• Caffeine commonly studied at ~2–6 mg/kg pre-exercise, with many using ~60 minutes pre-event. Half-life is typically hours, so late dosing disrupts sleep. ^{8 7}
• Beta-alanine typically 4–6 g/day for at least 2–4 weeks. Split doses can reduce paresthesia. ¹⁴
• Beetroot juice (nitrate) many studies use ~500 mL taken ~2.5–3 hours pre-exercise, providing roughly 310–682 mg nitrate depending on product. ⁷
• Psyllium meta-analyses for LDL show typical doses around ~10 g/day, often taken before meals, with benefits varying by dose and duration. ²⁰
• Melatonin effects on sleep onset are modest in meta-analysis. Dose and timing should match the problem. Many people do better with lower doses and correct timing rather than escalating. ^{1 2}

Start low and titrate: a concrete procedure

Use this whenever side effects are plausible (GI upset, sedation, stimulation).

• Days 1–3: 25–50% of target dose
• Days 4–7: 50–75% of target dose
• Week 2+: 100% of target dose if tolerated and still needed

Rules:

• Change only one of dose, timing, or formulation at a time.
• If side effects increase, revert to the last tolerable dose for 3–7 days.
• If you need to split dosing to tolerate it, treat that as a schedule change and hold everything else fixed.

Timing rules that prevent common failures

• Stimulants if a stimulant’s half-life is measured in hours, stop early enough that you still have a full sleep drive. Caffeine’s half-life is typically ~4–5 hours in ODS guidance, and individual variability is large. ⁷
• Sleep tools “bedtime dosing” is often wrong for circadian tools. For melatonin, timing relative to your desired sleep window matters. ²
• Fiber and GI tools start low, increase slowly, and anchor to meals to reduce variability. ¹⁹
• Fat-soluble nutrients take with a meal that contains fat to improve absorption. Vitamin D is fat-soluble and absorption is enhanced with concurrent fat in the gut. ¹¹

Safety and interactions that matter in practice

This section is not about being scared. It is about avoiding predictable mistakes that create regret.

Quality and contamination

• Supplements can have variable potency and may contain undeclared ingredients. Melatonin products, for example, have documented labeling and content issues in some analyses, and NCCIH highlights these concerns. ²
• If you are a tested athlete or your job has strict policies, contamination risk is not theoretical. The IOC consensus statement discusses both potential benefits and real risks from supplement use, including contamination and inadvertent doping. ²⁶

Practical quality rules:

• Prefer single-ingredient products over proprietary blends.
• Look for third-party testing (USP Verified, NSF Certified for Sport, Informed Sport) when the cost is justified by risk.
• Avoid “mega-dose” products unless you can explain why the dose is needed and you have a stop plan.

High-risk interaction classes and what to do

Minimal starter templates

These are examples. The rule still applies: build one-goal, one-stack, one change at a time. If you adopt a template with two ingredients, treat it as two separate experiments unless you have a strong reason to couple them (for example, theanine paired with caffeine to manage jitters).

Troubleshooting: what to do when stacks go wrong

“Nothing changed”

Most “no effect” outcomes are not failures. They are useful answers.

Run this checklist:

• Was the endpoint measurable and stable at baseline?
• Did you run the trial long enough for the compound’s likely time course?
• Did you change training, diet, caffeine, alcohol, or sleep schedule during the trial?
• Was the dose in a plausible range supported by human data?

If the protocol was sound and nothing moved, stop. Return to baseline for a week and test a different single lever.

“Too many variables”

If you added two things and it worked, you still do not know what worked.

Fix it by subtraction:

• Keep everything constant and remove one item for 7–14 days.
• If the benefit persists, the removed item was not necessary.
• If the benefit disappears, you have identified the active ingredient.

Side effects

Do not “patch” side effects with more supplements. Use the smallest change that addresses tolerability.

Common patterns:

• Sleep disruption caffeine too late, dose too high, or stimulant stacking. Reduce dose and move earlier. ⁷
• GI upset dose too high, formulation issue, or taking on an empty stomach when not tolerated. Split dose or take with meals, unless the protocol requires otherwise (for example, some peppermint oil trials use pre-meal dosing). ²¹
• Sedation or grogginess dose too high or timing too late. Reduce dose, move earlier, or stop.

Tolerance and “it stopped working”

This is most common with stimulants.

• For caffeine, consider using it selectively rather than daily if tolerance and sleep disruption are undermining the goal. ⁸
• Avoid escalating dose as the default. Escalation often buys short-term alertness at the cost of long-term sleep quality.

Withdrawal or rebound

If you stop something and feel worse, distinguish:

• Rebound (a transient swing past baseline)
• Loss of benefit (return to baseline)
• Withdrawal (physiologic adaptation, common with caffeine)

Plan exits:

• For caffeine, taper if intake is high and headaches or fatigue occur. ⁷
• For sedating agents, stop if adverse effects occur and reassess baseline sleep before adding a new variable.

Placebo and nocebo

Expectancy is part of the effect. You can reduce it without becoming obsessive.

• Use a consistent routine and do not “check for effects” every hour.
• If possible, use blinded capsules for single-ingredient tests.
• Anchor decisions to weekly averages, not single days.

In Unfair

The core problem with supplement stacks is not choice. It is workflow. Most people do not have a clean record of what they took, when they took it, and what changed.

Unfair supports the one-goal, one-stack method end to end:

• Stack templates Save a protocol as a named stack with dose, timing window, and notes on what it is trying to change.
• Reminders that match the protocol Not generic nudges. Reminders tied to the actual timing rules you set, including “pre-meal” or “pre-work” windows.
• Dose-event logging Log the dose at the moment it happens, including skipped doses. This matters more than perfect memory.
• Outcome tracking by goal Attach a primary endpoint to the stack and record quick daily metrics that roll up into weekly averages.
• Review loop A simple review view that compares baseline vs intervention windows so you can decide keep, adjust, or remove.
• Decision record Each change creates a written “why” so you do not retest the same failed idea six months later.

The goal is not to build the biggest stack. The goal is to build a small stack you can defend with your own data.

Continue with Supplement Foundations for Sustainable Results, Complete Guide to Supplement Stacks, Stacks vs Single Supplements.

Sources