Ashwagandha: What the RCTs Actually Show for Stress, Testosterone, and Sleep

Ashwagandha is the adaptogen that actually has data. Not decades of tradition and case reports. Actual double-blind placebo-controlled trials with standardized extracts on measurable endpoints — stress scales, cortisol, sleep latency, testosterone, anxiety.

Trials are small. Some have industry ties. A few raise replication concerns. Taken together, ashwagandha stands out among adaptogens as a compound with real signal, not just tradition.

This is the honest evidence walk-through. What it does. How much. In whom. And the safety caveats that get skipped in the marketing.

What ashwagandha is

Ashwagandha — Withania somnifera — is a small evergreen shrub in the nightshade family, native to India and the Middle East. Ayurvedic medicine has used the root for centuries. The name translates as "smell of the horse," referring to both the root's aroma and its traditional reputation for conferring stamina.

The active compounds are a class of steroidal lactones called withanolides. Withanolide A, withaferin A, and withanoside IV are the most studied. Withanolides are the reason the extract has to be standardized — raw root powder can vary widely in active content, and most published trials use extracts standardized to a specific withanolide percentage.

KSM-66 and Sensoril: the two standardized extracts that dominate the trial literature

Most published ashwagandha RCTs use one of two proprietary extracts. Knowing which is which matters when you read a study.

• KSM-66 (Ixoreal Biomed). Full-spectrum extract from ashwagandha root only. Standardized to at least 5 percent withanolides. Extraction is water-based rather than alcohol-based. The most widely used extract in published clinical trials — probably 30 to 40 percent of the modern literature.
• Sensoril (Natreon). Extract from ashwagandha root and leaf. Standardized to at least 10 percent withanolides. Higher withanolide content per gram, so effective doses are lower — 125 to 250 mg per day compared to KSM-66's typical 300 to 600 mg per day.

Unstandardized root powder — the cheap stuff on some shelves — has produced most of the underwhelming ashwagandha trials in the older literature. If a bottle does not say KSM-66, Sensoril, or a specific withanolide percentage, you are buying an unknown dose of active compound.

Stress and cortisol: the strongest evidence base

Stress reduction is the primary indication with the best RCT support. The trials are consistent enough that even skeptical reviewers acknowledge a real signal.

Chandrasekhar et al. 2012. 64 adults with chronic stress. KSM-66 300 mg twice daily for 60 days. Perceived Stress Scale scores dropped ~44 percent vs ~5 percent in placebo. Serum cortisol fell ~28 percent vs ~8 percent. The most-cited trial in the literature, with effect sizes large enough to warrant fair replication skepticism.

Salve et al. 2019. KSM-66 at 250 and 600 mg daily for 8 weeks in mild-to-moderate stress. Dose-dependent reductions in Perceived Stress Scale and Hamilton Anxiety. Cortisol down 14 to 22 percent. Smaller than Chandrasekhar, still real.

Lopresti et al. 2019. KSM-66 240 mg daily for 60 days in overweight adults. Significant reductions in DASS-21 stress and morning cortisol. Independent journal, cleaner design.

Auddy et al. 2008. Sensoril 125 to 500 mg daily for 60 days. Dose-dependent reduction in cortisol, C-reactive protein, and stress scores.

Pooling these: ashwagandha at standardized-extract doses reliably reduces perceived stress by 30 to 45 percent from baseline (versus small placebo movement) and cortisol by 15 to 25 percent. Effect sizes on stress scales run 0.7 to 1.0 SD in some trials — larger than most supplement categories manage.

Hold two thoughts. This is one of the better adaptogen evidence bases. And most trials come from Indian research groups with industry ties, in small samples, without long-term follow-up.

Sleep: modest signal, meaningful in some populations

Kelgane et al. 2020. 50 older adults with sleep complaints. KSM-66 600 mg daily for 12 weeks. Improvements in sleep quality, sleep onset latency, and total sleep time on subjective measures.

Langade et al. 2019 and 2021. Two trials in adults with insomnia. KSM-66 300 mg twice daily for 8 to 10 weeks. Reductions in sleep-onset latency of roughly 20 to 30 minutes and improved sleep-efficiency scores.

Deshpande et al. 2020. 150 adults with disturbed sleep. KSM-66 120 mg daily for 6 weeks. Modest subjective sleep improvement.

The sleep signal is real but smaller in magnitude than the stress signal. Ashwagandha is not a sedative. It appears to help sleep primarily by reducing pre-bed anxious arousal — the mechanism route is stress reduction, not direct hypnotic action. If your sleep problem is racing thoughts and rumination at bedtime, ashwagandha has a plausible mechanism. If your problem is fragmented mid-night waking or physical restlessness, the mechanism fits less cleanly.

Testosterone: modest signal in specific populations

This is where the marketing runs ahead of the data. Ashwagandha is sold in the men's health category as if it were a testosterone booster. The trials are more nuanced.

Lopresti et al. 2019. The same overweight-adult trial mentioned under stress also measured hormones. KSM-66 240 mg daily for 60 days produced a ~15 percent increase in serum testosterone in overweight men versus placebo. Effect was modest and in a specific population.

Wankhede et al. 2015. 57 young men on a resistance-training program. KSM-66 300 mg twice daily for 8 weeks. Testosterone rose by ~15 percent versus placebo. Also improved strength and lean mass gains from training.

Ambiye et al. 2013. 46 men with low sperm count. KSM-66 675 mg daily for 90 days. Improved sperm concentration and motility with modest testosterone rise.

The pattern: ashwagandha appears to modestly raise testosterone in men who are stressed, overweight, low-fertility, or resistance-training. Effect size is roughly 10 to 20 percent — meaningful biologically, small compared to actual testosterone-replacement therapy, and inconsistent in men with already-normal testosterone. It is not a substitute for medical evaluation of hypogonadism.

Female hormonal effects are much less well-studied. A few trials have looked at sexual function in women with mixed results. Not enough data to draw strong conclusions.

Anxiety: multiple modest trials

Related to but distinct from stress. Anxiety trials have been small and use varied scales.

Andrade et al. 2000 in generalized anxiety disorder patients showed improvement on Hamilton Anxiety Scale over 6 weeks versus placebo. Various KSM-66 trials show anxiety scales moving alongside stress scales, though anxiety is rarely the primary endpoint.

Ashwagandha may reasonably be considered for mild-to-moderate anxiety. It is not a replacement for pharmacotherapy in moderate-to-severe GAD, and it should not be stacked on benzodiazepines or SSRIs without clinical supervision — additive sedation and interactions are poorly characterized.

Dose, timing, and duration

Practical protocol from the trial literature:

• KSM-66: 300 to 600 mg per day, typically split into two doses.
• Sensoril: 125 to 250 mg per day, once or twice daily.
• Timing: Morning and evening dose. Ashwagandha is not sedating in the acute-dose sense, but the stress-reducing effect may be most useful before bed.
• Duration: Effects on stress scales appear within 4 to 6 weeks; cortisol and hormonal effects mature over 8 to 12 weeks. Trials shorter than 6 weeks tend to under-detect the effect. Do not judge at two weeks.
• With or without food. With food is easier on the stomach. Some users report mild nausea on an empty stomach.

Safety: what the marketing skips

Ashwagandha is generally well-tolerated. Common side effects are mild GI upset, drowsiness, and occasional headache. Rare but real signals worth knowing:

Hepatotoxicity case reports. Post-2019, a small but growing number of case reports have described drug-induced liver injury attributed to ashwagandha (Björnsson et al. 2020 Iceland surveillance data and follow-up American case series). Typical presentation is cholestatic hepatitis 2 to 12 weeks after starting, usually reversible on discontinuation. Incidence is rare — probably less than one per thousand users — but not zero. Anyone with existing liver disease should not use ashwagandha without clinician supervision. If you develop yellow eyes, dark urine, or right-upper-quadrant pain during a course, stop and get liver enzymes checked.

Thyroid effects. Ashwagandha can raise thyroid hormone levels (T3 and T4) modestly. This has been shown in small trials — Sharma et al. 2018 in subclinical hypothyroid patients showed meaningful T3 and T4 increases and TSH decrease over 8 weeks. The clinical implication cuts two ways:

• People with subclinical hypothyroidism or Hashimoto's may see mild improvement on ashwagandha.
• People with hyperthyroidism or Graves' disease should avoid it — additional thyroid stimulation is exactly the wrong direction.
• People on thyroid replacement (levothyroxine) may need dose adjustments if adding ashwagandha; get labs before and after.

Autoimmune concerns. Immune-modulating effects are theoretical. Anyone with active autoimmune disease should discuss with their rheumatologist or endocrinologist before starting.

Pregnancy. Traditionally contraindicated in pregnancy due to potential effects on smooth muscle. Modern safety data in pregnancy is essentially absent. Avoid.

Sedative and psychiatric medications. Additive sedation with benzodiazepines and possibly SSRIs; interactions with lithium have been proposed. Do not layer ashwagandha on psychiatric drugs without your prescriber.

The cycling question

A common recommendation is to "cycle" ashwagandha — take it for 8 to 12 weeks, then stop for 2 to 4 weeks, then repeat. The rationale is usually stated as tolerance prevention or adrenal recovery.

The evidence for cycling being necessary is weak. No published trials have shown tolerance to ashwagandha's stress or hormonal effects over the trial durations used (up to 12 to 16 weeks). No published trials have compared continuous versus cycled dosing in a controlled way. The "adrenal fatigue" framework the cycling recommendation rests on is not a recognized medical concept — cortisol dysregulation is a real thing, but it is not treated by cycling herbs.

A defensible middle path: take ashwagandha continuously through a stress-heavy period (a demanding project, a training block, a life transition), and stop when the stressor resolves. If you notice diminishing benefit after several months, take a two-week break and reassess. But there is no biological argument for a rigid 8-on-2-off schedule.

What ashwagandha is not

Managing expectations is half of getting the actual benefit.

• Not a stimulant. Does not increase energy acutely. Any "more energy" people report is downstream of better sleep and lower baseline anxiety, over weeks.
• Not a sedative. Does not knock you out. Sleep benefit comes from reducing pre-sleep arousal, not from direct hypnotic action.
• Not a testosterone-replacement equivalent. The magnitude of effect on T is 10 to 20 percent in specific populations. TRT moves T by hundreds of ng/dL.
• Not a treatment for major depression, anxiety disorder, PTSD, or clinical burnout. Mild-to-moderate stress support, in generally healthy people, with room for adjunctive use under clinical supervision.
• Not a general-purpose "adaptogen." The word adaptogen is a marketing category, not a pharmacological one. Ashwagandha does specific things via specific pathways. Buy it for those things, not for the vibe.

A 60-day ashwagandha trial

The literature suggests 6 to 8 weeks is the minimum honest test window. Sixty days is a good balance between waiting long enough to see the effect and not committing to a year.

1. Choose a standardized extract. KSM-66 or Sensoril, labeled clearly. Skip generic root powder.
2. Baseline for one week. Track subjective stress on a 1 to 10 scale each morning. Track sleep-onset latency. Optionally, get a baseline morning cortisol via a saliva test kit or a fasted serum cortisol if you have upcoming labs anyway.
3. Start at KSM-66 300 mg once daily with dinner. Add a second 300 mg dose in the morning after 3 to 5 days if tolerating well. Or use Sensoril at 125 to 250 mg daily.
4. Continue for 60 days. Do not change other variables — no simultaneous new supplements, no new medications, no starting-and-stopping caffeine.
5. Evaluate at day 60. Reasonable expectations: subjective stress score down 2 to 3 points from baseline. Sleep-onset latency down 10 to 20 minutes. Optional cortisol retest, down 15 to 25 percent. Testosterone up 10 to 20 percent if male and specifically overweight or stressed at baseline.
6. Decide. If the numbers moved, continue. If they did not, stop — the compound is not the bottleneck for your particular biology.

The honest summary

Ashwagandha earns a spot in the small group of supplements with real, replicated human-trial evidence. Stress and cortisol effects are meaningful. Sleep signal is modest and mediated through stress reduction. Testosterone effects are real in specific populations and modest in size. Anxiety benefits are consistent with the stress-reduction mechanism.

Buy a standardized extract. Take the trial-supported dose for 8 to 12 weeks. Watch for liver signals and thyroid-related contraindications. Do not stack it on psychiatric medications without your prescriber. Do not expect it to fix a life crisis — it is a modulator, not a rescue drug.

And if the trial does not move your numbers, stop. That is how you tell the difference between working supplements and expensive placebo.