Natural Testosterone Optimization: What Actually Works (and What Doesn't)

Key Takeaways

Search for "how to raise testosterone naturally" and the answer comes back as a shopping list. The trial literature answers a different question — how much does each lever actually move the number? — and when the levers are sorted by that, the list inverts. The interventions with the largest measured effects cost nothing and come in no bottle. The interventions that come in bottles produce small effects, and only some of them produce any effect at all.

This is that sorted list. Four tiers, each with the effect size the papers actually report, and a closing section on the point at which the levers stop being the answer.

The Number You Are Trying to Move

Testosterone declines about 1–2% a year from the thirties, which means a "low" reading in a 45-year-old is usually a mixture of age, sleep, body fat and measurement noise rather than disease. In the European Male Ageing Study — a random population sample of 3,369 men aged 40–79 — sexual symptoms became discernibly more probable as total testosterone fell through the 8.0–13.0 nmol/L range, and late-onset hypogonadism as a syndrome required at least three sexual symptoms plus a total testosterone below 11 nmol/L and free testosterone below 220 pmol/L.

That definition matters for reading everything below: most of the effects reported here are in the range of 1–5 nmol/L. For a man at 9 nmol/L, that is the difference between a diagnosis and not. For a man at 20, it is noise.

Tier 1 — The Levers That Move Testosterone the Most

Sleep

The cleanest demonstration of a lifestyle effect on testosterone is a small, tightly controlled University of Chicago study. Ten healthy men aged around 24 spent three nights with ten-hour bedtimes, then eight nights restricted to five hours. Blood was sampled every 15–30 minutes for 24 hours in each condition.

After one week of restriction, daytime testosterone fell from 18.4 to 16.5 nmol/L (P = 0.049), with the drop most pronounced between 2 PM and 10 PM (17.9 to 15.5 nmol/L, P = 0.02). The authors put it at a 10–15% decrease — and pointed out that normal ageing costs 1–2% per year, so a week of short sleep produced the testosterone signature of a decade of ageing. Cortisol did not change; vigour scores fell steadily across the week.

Ten men is a small sample and the sleep restriction was severe. But it was also a condition experienced by at least 15% of the working population, and the mechanism is not controversial: most daily testosterone release happens during sleep. No supplement in Tier 3 has an effect size that would survive a five-hour sleep schedule.

Body fat

Obesity produces a specific, reversible form of low testosterone — hypogonadotropic, meaning the pituitary signal falls alongside the hormone. A meta-analysis of 24 studies asked whether losing the weight reverses it.

It does. Both low-calorie diet and bariatric surgery produced significant rises in total testosterone (P < 0.0001). The size depended on the method: 2.87 nmol/L (95% CI 1.68–4.07) for diet and 8.73 nmol/L (6.51–10.95) for surgery. The rise was larger in men who lost more weight, in younger men, and in non-diabetic men with greater baseline obesity. Weight loss also lowered oestradiol and raised gonadotropins, which is what reversal of the mechanism should look like.

For context, an 8.73 nmol/L rise is roughly the entire distance across the EMAS "symptoms become likely" zone. It is also the one Tier 1 lever that serves fertility as well as testosterone, since obesity independently raises the odds of a low sperm count.

Alcohol

This one belongs in Tier 1 with a caveat about how much. A 2024 meta-analysis of 17 observational studies found no significant testosterone difference between alcohol consumers and non-consumers overall (SMD 0.03, 95% CI −0.11 to 0.16); the significant, damaging changes were concentrated in alcohol use disorder, with feedback changes in LH. Heavy chronic drinking suppresses the axis. A moderate intake does not show up in the pooled data.

The practical version: if drinking is heavy or daily, it is a Tier 1 lever. If it is occasional, it is not the problem.

Tier 2 — Correcting What Is Wrong

Vitamin D — when deficient

The trial usually cited for vitamin D and testosterone was a one-year analysis of 54 overweight men on a weight-reduction programme, randomised to 3,332 IU (83 μg) daily or placebo. Both groups started deficient (25(OH)D below 50 nmol/L) and at the low end of the testosterone reference range.

In the vitamin D group, total testosterone rose from 10.7 to 13.4 nmol/L (P < 0.001), bioactive testosterone from 5.21 to 6.25 (P = 0.001), and free testosterone from 0.222 to 0.267 nmol/L (P = 0.001). Placebo showed no significant change.

That is a meaningful effect — about 25% — but the population defines it. These men were deficient. A 2024 systematic review of 52 "testosterone booster" studies found ten trials of cholecalciferol and concluded that, like most boosters, it fails to raise total testosterone in the populations studied broadly. Vitamin D corrects a deficiency. It is not a testosterone supplement for the replete. Dose to a level, as set out in the vitamin D dosing guide.

Zinc — when deficient

The same logic, with older data. In a series of experiments from Wayne State, cellular zinc correlated with serum testosterone across 40 healthy men (lymphocyte zinc, r = 0.43, P = 0.006). Four young men put on a zinc-restricted diet for 20 weeks saw testosterone fall from 39.9 to 10.6 nmol/L (P = 0.005). And nine marginally zinc-deficient elderly men supplemented for six months rose from 8.3 to 16.0 nmol/L (P = 0.02).

Nine men is a very small sample; the effect is very large. It is also an effect of repletion. No trial has shown zinc raising testosterone in a man whose zinc is normal, and the 2024 systematic review of boosters found the five zinc/magnesium trials unpersuasive as a class. Measure serum zinc, supplement if low, and stop treating it as a default.

Resistance training — honestly

Resistance training belongs on every list for body composition, insulin sensitivity and the strength that keeps older men independent. It does not belong on a list of things that raise resting testosterone, and the meta-analysis is unambiguous.

Eleven randomised trials, 421 insufficiently active but apparently healthy men aged 19–75, aerobic, resistance or combined training for a median of 12 weeks: the pooled effect on resting total testosterone was SMD 0.00 (95% CI −0.20 to 0.20). Subgroup analysis found no difference by training mode, age, body mass, or testosterone measure.

The testosterone benefit of training runs through Tier 1 — through fat loss — not through the session itself. Acute post-exercise spikes are real and short-lived, and they are not what a morning blood test measures.

Tier 3 — Supplements With Human Randomised Evidence

A useful frame comes from the 2024 systematic review in International Journal of Impotence Research, which went through 52 studies of 27 proposed "testosterone boosters." Its finding: most fail to increase total testosterone. The exceptions it named as possibly effective included Tongkat Ali and ashwagandha for healthy men, and Tongkat Ali and purified shilajit for men with late-onset hypogonadism. "Possibly effective" is the right register for this whole tier.

Ashwagandha

The best-designed trial is a 16-week randomised, double-blind, placebo-controlled crossover in overweight men aged 40–70 with mild fatigue, using a standardised extract delivering 21 mg of withanolide glycosides daily for 8 weeks per arm. Fifty-seven men enrolled; 43 completed all 16 weeks.

Ashwagandha produced an 18% greater increase in DHEA-S (P = 0.005) and a 14.7% greater increase in salivary testosterone (P = 0.010) than placebo. Cortisol and oestradiol did not differ. And then the finding that should sit next to the headline: fatigue, vigour, and sexual and psychological well-being improved over time in both groups with no significant between-group difference. The hormone moved; nothing the men could feel moved with it.

That is a real, modest, biochemical effect from a single well-run trial. The existing profile at ashwagandha covers the cortisol literature, which is broader.

Tongkat Ali

Tongkat Ali (Eurycoma longifolia) has the largest evidence base in this tier, mostly for free testosterone via SHBG reduction. It has already been reviewed in detail in Tongkat Ali and Fadogia Agrestis: The Evidence-Based Guide and this article does not re-litigate it. The 2024 systematic review counted three trials and rated it possibly effective in both healthy and hypogonadal men — the strongest rating any botanical received. Where it belongs, with boron and zinc, is the Hormonal Optimization Stack.

Shilajit

Purified shilajit has one placebo-controlled testosterone trial: healthy men aged 45–55, 250 mg twice daily for 90 days, with total testosterone, free testosterone and DHEAS significantly higher than placebo (P < 0.05) and LH and FSH maintained. The abstract reports no effect sizes, and both shilajit trials come from a single Kolkata institution using one manufacturer's product. Its more interesting data are in sperm parameters, not testosterone — which is why it sits in the Male Vitality Protocol rather than the testosterone stack. The full profile is at shilajit.

Tier 4 — What Does Not Work in Humans

Tribulus terrestris

Tribulus is the most-sold "testosterone" herb and the trial evidence against it is direct. Twenty-one healthy men aged 20–36 were randomised to Tribulus extract at 10 or 20 mg/kg/day or placebo for four weeks, with testosterone, androstenedione and LH measured at six time points. No significant difference on any hormone (all P > 0.05): testosterone 15.75 and 16.32 nmol/L on the two doses against 17.74 on placebo. The authors concluded that Tribulus saponins possess neither direct nor indirect androgen-increasing properties. The 2024 booster review, with four Tribulus trials in hand, agreed.

D-aspartic acid

Worse than null. Twenty-four resistance-trained men were randomised to 3 g/day, 6 g/day, or placebo for two weeks. At 6 g/day, total testosterone was significantly reduced (P = 0.03) and free testosterone was significantly lower than placebo (P = 0.005). The 3 g dose did nothing. An earlier claim of benefit came from sedentary men; in trained men the compound is inert at the low dose and suppressive at the high one.

Fadogia agrestis

Fadogia has rodent data and no human clinical trial, along with dose-dependent testicular toxicity signals in the animal studies. It cannot be placed in any evidence tier because there is no human evidence to place. The Fadogia guide sets out what the animal literature does and does not show.

Expected Effect Size by Lever

Two things stand out. Every effect above about 3 nmol/L comes from correcting something — obesity, deficiency, sleep debt. And every supplement effect in a eugonadal population is small enough to be erased by a bad week of sleep.

Putting the Tiers in Order

1. Sleep seven to eight hours and treat sleep apnoea if it exists. This is the cheapest and fastest lever, and the only one with an effect measured in days.
2. Lose body fat if BMI is over 30. The largest sustained effect in the literature, and the only one that also serves fertility.
3. Stop heavy drinking. Moderate intake is not the issue; alcohol use disorder is.
4. Measure 25(OH)D and zinc; correct what is low. Do not supplement what is normal.
5. Train for body composition, expecting nothing from the session itself for resting testosterone.
6. Then, and only then, consider Tier 3 — ashwagandha, Tongkat Ali — for a modest additional effect, in the sequence set out in the Hormonal Optimization Stack.
7. Skip Tier 4 entirely.

A man who does steps 1–5 and still has symptoms and a confirmed low reading has moved past what any of this can do.

When Numbers, Not Habits, Are the Problem

Everything above assumes the low reading is the product of something reversible. Sometimes it is not. Primary testicular failure, pituitary disease, and a genuine, persistent hypogonadism that survives good sleep, normal body fat and corrected deficiencies are medical conditions, and no supplement tier addresses them.

The line between "optimise the levers" and "this is a diagnosis" is drawn by two fasting morning measurements, a symptom pattern, and — where the total is borderline — a proper free testosterone. What that evaluation involves, what testosterone therapy has and has not been shown to do in randomised trials, and why it is the wrong tool for a man who wants children, is the subject of When Lifestyle Isn't Enough: The TRT Decision Framework.

Frequently Asked Questions

How quickly does sleep affect testosterone?

Within a week. The restriction study measured a 10–15% daytime decrease after seven nights of five-hour sleep. Recovery has not been formally timed in the same design, but the mechanism — testosterone release during sleep — suggests it is equally fast in the other direction.

I lift weights four times a week. Why is my testosterone not higher?

Because resistance training does not raise resting testosterone in men with normal levels. The pooled effect across 11 randomised trials was zero. Training improves body composition, and fat loss improves testosterone — but a lean man who trains hard should not expect the training itself to add to the number.

Should I take vitamin D and zinc anyway, just in case?

No. The trials that show an effect enrolled deficient men. A blood test for 25(OH)D and serum zinc costs less than a month of supplements and tells you whether Tier 2 applies to you at all.

Is ashwagandha worth taking for testosterone?

It produced a 14.7% greater rise than placebo in one well-designed trial — and no difference in how the men felt. If the goal is a biochemical change, it has evidence. If the goal is energy or libido, that trial did not find it. Its cortisol evidence is broader and may be the better reason to take it.

What about Tongkat Ali and boron together?

That is the SHBG-reduction route, covered in the Boron + Tongkat Ali protocol guide. It targets free rather than total testosterone and is the most evidence-backed supplement combination for men whose goal is the hormone number. It is a separate route from fertility.

My testosterone is low and none of this has changed it. What now?

Two fasting morning tests, both low, plus symptoms, is the threshold at which this becomes a clinical question. The next article in this series sets out what a hypogonadism diagnosis requires and what testosterone therapy does and does not do.

Related Research

• When Lifestyle Isn't Enough: The TRT Decision Framework
• The Male Vitality Protocol: Fertility and Sexual Function Beyond Testosterone
• Tongkat Ali and Fadogia Agrestis: The Evidence-Based Guide to Natural TRT Support
• Hormonal Optimization Stack
• Shilajit (Purified)