Ashwagandha (Withania somnifera): the AU general practice evidence appraisal

What this consultation is usually about

Ashwagandha is one of the most common supplement questions in AU general practice in 2026. It arrives via the Huberman Lab podcast, social media adaptogen discourse, and AU women’s-wellness marketing for stress, fatigue and hormonal balance. The clinical question the GP actually needs to answer is: does this patient’s intended use have a plausible evidence base, and is it safe for them specifically?

The short framework: the TGA lists ashwagandha as a low-risk complementary medicine under the AUST L scheme, which confirms it has passed a basic safety threshold — not that efficacy has been reviewed. The 2025 TGA safety advisory on hepatotoxicity has shifted the risk-benefit calculus for higher-risk patients. For most healthy adults asking about stress and sleep, the evidence supports a shared decision-making conversation rather than a blanket recommendation in either direction.

A. Core clinical — the AU general practice framework

Botanical identity and standardisation

Withania somnifera (Solanaceae — the nightshade family) is a small woody shrub native to India, the Middle East and parts of Africa, used in Ayurvedic medicine for approximately 3,000 years. The active phytochemicals are withanolides — steroidal lactones concentrated in the root — alongside withanosides, sitoindosides, alkaloids and iron.

Standardisation matters critically. Most retail “ashwagandha” capsules are uncharacterised root powder with negligible withanolide content. The two extracts that have generated virtually the entire clinical evidence base are:

• KSM-66 (Ixoreal Biomed) — root only, ≥5% withanolides by HPLC — used in the majority of published RCTs at 300 mg twice daily (= 600 mg/day).
• Sensoril (Natreon) — root and leaf, ≥10% withanolide glycosides — studied at 240–250 mg once daily.

A third extract, Shoden (Arjuna Natural), is higher-potency (≥35% withanolide glycosides) at smaller doses (60–120 mg/day), with emerging trial data. Generic powders are not equivalent to any of these.

Proposed mechanisms

The evidence-supported mechanisms include:

• HPA-axis modulation — chronic withanolide dosing attenuates the ACTH-cortisol response to stress in animal models; human RCTs show ~20–30% morning serum cortisol reduction in stressed populations.
• GABAergic activity — sitoindosides and selected withanolides show GABA-A receptor binding affinity in vitro, plausibly underpinning the anxiolytic and sleep-onset signal.
• Modest thyrotropic effect — Sharma 2018 demonstrated TSH reduction of ~17% and T4/T3 elevation in subclinical hypothyroid patients on 600 mg KSM-66 for 8 weeks — clinically important in both directions (see Special Populations).
• Testosterone signal — Lopresti 2019 reported ~14% serum testosterone rise in middle-aged men at 8 weeks; Wankhede 2015 showed larger testosterone and strength gains in resistance-training men. Both trials carry significant manufacturer-funding bias.

Dose and duration guidance

Indication	Extract	Dose	Duration
Stress / anxiety	KSM-66	300 mg BD with food	8 weeks; review and stop-or-continue
Stress / anxiety	Sensoril	240 mg once daily (evening meal)	60 days
Sleep onset	KSM-66	300 mg once daily 60–90 min before bed	6–8 weeks
Maximum continuous use (monitoring)	Any	As above	12 weeks, then LFT recheck + stop-and-review

Onset of perceived effect is typically 2–4 weeks; peak at 8 weeks. Mega-dosing above 1,200 mg/day standardised extract has no evidence of greater benefit and increases hepatotoxicity and immune-stimulation risk.

B. Evidence appraisal — what the trials actually show

Cortisol and perceived stress

Chandrasekhar 2012 (n=64, 300 mg KSM-66 BD, 60 days) showed Perceived Stress Scale scores down 44% and morning serum cortisol down 28% versus placebo in stressed Indian adults. Salve 2019 (n=60, 250 mg and 600 mg KSM-66) showed dose-dependent cortisol reductions of 14–28%. Lopresti 2019 (n=60, 240 mg Sensoril) showed morning cortisol down ~23% and significant Perceived Stress Scale improvement.

The 2025 Della Porta meta-analysis (12 RCTs, n=1,002) confirms a pooled cortisol-lowering effect — but with heterogeneity (I² >70%) and wide confidence intervals. Structural weaknesses in the evidence base include: ~80% of trials conducted in single-centre Indian populations; dominant manufacturer funding (Ixoreal, Natreon); small sample sizes (median n ≈ 60); cortisol measured as a single morning serum sample rather than diurnal AUC or cortisol awakening response; and no trial comparing ashwagandha against an active comparator (SSRI, CBT, melatonin).

Anxiety

The anxiety signal mirrors the cortisol signal in the same trials. Chandrasekhar 2012 showed Hamilton Anxiety Rating Scale reduction of 56% versus placebo — a large effect size unlikely to survive in a larger, independent Western academic population. Pratte 2014 systematic review of 5 RCTs found all favoured ashwagandha for anxiety versus placebo. The Della Porta 2025 pooled analysis showed SMD −1.13 (95% CI −1.85 to −0.41) for anxiety scales — plausible but heavily influenced by manufacturer-funded trials.

Clinical bottom line: moderate evidence for anxiety symptom reduction in mild–moderate chronically stressed populations; no evidence of equivalence to first-line treatments (SSRIs, CBT); reasonable adjunct in shared decision-making.

Sleep

Langade 2019 (n=80, 300 mg KSM-66 BD, 10 weeks) showed sleep onset latency reduced by 40%, total sleep time increased 12%, and sleep quality improved by 39% on the Pittsburgh Sleep Quality Index versus placebo. These are compelling numbers — from a small single-centre trial using subjective measures only (no actigraphy, no polysomnography), funded by Ixoreal.

The sleep effect is plausibly downstream of the anxiolytic GABAergic action rather than a primary hypnotic mechanism. Ashwagandha for sleep is best positioned as an adjunct after CBT-I and sleep hygiene have been addressed (see the insomnia management framework).

C. Safety profile — the hepatotoxicity signal

What the evidence shows

Björnsson et al. 2020 (Liver International) described 5 patients (3 Iceland, 2 US Drug-Induced Liver Injury Network) with biopsy-confirmed cholestatic-hepatocellular DILI. RUCAM causality scoring was “probable” to “highly probable”. Latency was 2–12 weeks; all recovered fully on discontinuation. This triggered Nordic regulatory tightening: Iceland restricted ashwagandha sales in 2023 (Lyfjastofnun); Denmark moved it to pharmacy-only; Sweden and Finland issued consumer advisories. The pattern informed the 2025 TGA safety advisory.

The 2025 TGA safety advisory flagged accumulating Australian case reports, tightened pharmacovigilance obligations on sponsors, and recommended consumer awareness. No TGA-confirmed fatalities appear in published series, but case reports include acute liver failure listings in international pharmacovigilance databases.

The signal is idiosyncratic, not dose-related — it appears across products, brands and standardisations, without a clear “safe” formulation identified.

Monitoring protocol

For any patient choosing to use ashwagandha after shared decision-making:

1. Baseline LFTs (ALT, AST, ALP, GGT, bilirubin) before starting in patients with risk factors: pre-existing liver disease, alcohol intake >2 standard drinks/day, hepatotoxic co-medications, age >60, obesity or metabolic dysfunction-associated steatotic liver disease.
2. Recheck LFTs at 6–8 weeks and at 12 weeks for continuous users.
3. Stop immediately and seek GP review for: jaundice, scleral icterus, dark urine, pale stool, right upper quadrant pain, persistent nausea or unusual fatigue.
4. Hepatotoxic co-medications to flag — paracetamol >3 g/day chronically, statins, isoniazid, methotrexate, valproate, amoxicillin-clavulanate, nitrofurantoin, ketoconazole, and traditional herbal products including kava, comfrey and pyrrolizidine-alkaloid-containing herbs.
5. Report adverse events to the TGA Adverse Event Reporting Portal.

D. Australian operations

TGA and market landscape

Ashwagandha products sold over the counter must carry an AUST L number under the TGA listed-medicines scheme. Permitted indications are low-level traditional claims (“traditionally used in Ayurvedic medicine to support stress and sleep”) — not structure/function claims, which require an AUST R registration.

Common Australian brands include Blackmores Stress + Adrenal Support, Caruso’s Stress + Anxiety, Fusion Stress and Anxiety, Herbs of Gold Ashwagandha 6500, and numerous private-label and online-only products. Retail pricing runs AUD 25–60/month — KSM-66 and Sensoril-standardised products command the higher end; generic root powder the lower. The AUST L number does not differentiate these.

Pharmacist counselling is not mandatory but increasingly common as pharmacy chains adopt voluntary risk-counselling protocols mirroring the EU regulatory posture.

MBS billing

The consult bills under standard MBS items 23/36/44. If anxiety or adjustment disorder is the primary clinical diagnosis:

• Mental Health Care Plan — items 2715/2717 (MHST GPs) for Better Access psychology referral; review item 2712.
• GP Chronic Disease Management Plan — items 965/967 if chronic anxiety meets comorbid criteria.
• LFTs — MBS item 66512 for monitoring.
• TFTs — MBS item 66716 for thyroid monitoring in at-risk patients.

No PBS subsidy exists for ashwagandha — entirely patient-self-funded.

Documentation prompts for shared decision-making

• Document shared decision-making: patient understands evidence is moderate, ashwagandha is adjunctive to first-line care, and a clear stop-and-review plan is in place.
• Document contraindication screening: pregnancy, breastfeeding, autoimmune disease, hyperthyroidism, levothyroxine use, pre-existing liver disease, hepatotoxic co-medications.
• Document the specific product, extract type (KSM-66 / Sensoril / generic) and dose.
• Document baseline LFTs and the planned 6–8-week recheck.
• Document safety-netting: stop immediately and seek review for jaundice, dark urine, right upper abdominal pain, fatigue or nausea; symptoms suggesting thyroid or autoimmune flare also warrant early review.

E. Special populations

Pregnancy and breastfeeding. Absolute contraindication. Abortifacient in Ayurvedic traditional classification; uterine-stimulant in animal models; no human safety data. RANZCOG and NPS MedicineWise both advise against unregulated herbal use in pregnancy and breastfeeding.

Thyroid disease. Do not use without monitoring in any patient with known thyroid disease. The modest thyrotropic effect (TSH ↓ ~17%, T4/T3 rise) documented in subclinical hypothyroid patients creates risk of tipping patients with Graves’ disease, autonomously functioning nodules or borderline thyroid status into overt hyperthyroidism. Patients on levothyroxine require TFT recheck at 6–8 weeks if they choose to use ashwagandha.

Active autoimmune disease. Avoid in Hashimoto’s thyroiditis, lupus, rheumatoid arthritis or multiple sclerosis. Ashwagandha has Th1 immune-stimulating properties in vitro — theoretical risk of disease flare in active autoimmune conditions.

Pre-existing liver disease. Avoid. The idiosyncratic hepatotoxicity signal is additive with pre-existing hepatic vulnerability. The TGA 2025 advisory specifically flags this group.

Athletes. Ashwagandha is not currently on the WADA 2026 Prohibited List and is permitted in and out of competition. However, competitive athletes using unregulated supplements face strict liability for any prohibited substance contamination. Advise athletes in tested sports to use only third-party batch-tested products (HASTA in Australia; Informed Sport internationally) and to declare supplement use to team medical staff.

PCOS. The modest testosterone-raising signal in men (Lopresti 2019; Wankhede 2015) carries theoretical risk of worsening androgenic features — hirsutism, acne — in women with polycystic ovary syndrome. Not recommended in PCOS without specialist input.

Older adults. The sedative-adjacent GABAergic mechanism creates additive risk with benzodiazepines, Z-drugs and opioids — drug interactions relevant in polypharmacy-heavy older patients. CYP3A4 metabolism creates theoretical interactions with clarithromycin, ketoconazole and carbamazepine.

When to escalate

Refer or escalate when:

• Liver injury suspected — jaundice, dark urine, RUQ pain, raised transaminases — for urgent hepatology review; report to the TGA adverse event portal.
• Thyroid symptoms or TFT derangement in a patient using ashwagandha — for endocrinology review; stop ashwagandha immediately.
• Autoimmune flare in a patient using ashwagandha — for rheumatology or relevant specialist review.
• Anxiety disorder requiring formal management — do not allow supplement focus to delay MHCP referral and Better Access psychology.
• Pregnancy discovered while using ashwagandha — stop immediately; obstetrics review; reassurance that the latency-based hepatotoxicity pattern makes fetal risk from short-term exposure a theoretical rather than documented concern, but specialist input appropriate.

What this article is and is not

This is general health information drawn from peer-reviewed evidence and current Australian regulatory guidance — the 2025 TGA safety advisory, NPS MedicineWise consumer information, RANZCOG herbal-medicines-in-pregnancy position, and the primary RCT and meta-analysis literature. It is not personal medical advice and does not create a doctor–patient relationship. Decisions about whether to use complementary medicines, including ashwagandha, are shared decisions made with your own GP, who can assess your specific circumstances, contraindications and medications.

For Australian consumer information: HealthDirect — Complementary and alternative medicines, Better Health Channel — Herbal medicine, NPS MedicineWise — Herbal medicines.

If you are experiencing liver symptoms (jaundice, dark urine, abdominal pain) after using any supplement, seek urgent GP or emergency department review.

---

Sources cited

1. TGA 2025 — Ashwagandha and reports of liver injury
2. NPS MedicineWise — Herbal medicines
3. RANZCOG — Herbal medicines in pregnancy
4. HealthDirect — Complementary and alternative medicines
5. Better Health Channel — Herbal medicine
6. Chandrasekhar K et al. 2012 — KSM-66 stress and anxiety RCT
7. Lopresti AL et al. 2019 — Sensoril stress and cortisol RCT
8. Salve J et al. 2019 — KSM-66 adaptogenic and anxiolytic RCT
9. Langade D et al. 2019 — KSM-66 insomnia and anxiety RCT
10. Della Porta M et al. 2025 — Ashwagandha meta-analysis (PMC12242034)
11. Björnsson ES et al. 2020 — Ashwagandha-induced liver injury case series
12. Sharma AK et al. 2018 — KSM-66 subclinical hypothyroidism RCT
13. Lyfjastofnun (Iceland Medicines Agency)
14. WADA 2026 Prohibited List
15. TGA Listed Medicines overview