N-Acetylcysteine (NAC)

The unambiguous, established use — a hospital protocol, not a self-care one. A toxic paracetamol metabolite (NAPQI) depletes hepatic glutathione and attacks liver protein; NAC refills the glutathione pool fast enough to prevent hepatocyte death, and it is standard of care given intravenously or orally, dosed by body weight and timed against ingestion. A Cochrane review identified 11 randomised trials (700 participants) across all overdose interventions but could pool almost none, and rated the evidence low to very low quality — the trials are old and underpowered, because withholding an antidote that clearly works is not ethical ^{1Reference 1Chiew AL et al. · 2018Systematic reviewInterventions for paracetamol (acetaminophen) overdose — Cochrane systematic reviewView study →}.

Its status as first-line treatment rests on decades of clinical experience, not on a single decisive RCT.

Gap: the strong recommendation is clinical consensus and mechanism, not high-quality trial data — and none of this transfers to over-the-counter supplement use. Suspected overdose is a medical emergency.

NAC’s original indication. As a mucolytic its free thiol group breaks the disulfide bonds that make mucus viscous — a direct chemical effect, separate from the glutathione story. A meta-analysis of 11 trials (775 on NAC against 789 on placebo) found oral NAC reduced chronic-bronchitis exacerbations (RR 0.81, 95% CI 0.69–0.93) and improved symptoms (RR 1.68, 1.13–2.52), with adverse events no different from placebo; the effective pattern was a lower dose, under 1,200 mg/day, taken for at least three months ^{2Reference 2Wei J et al. · 2019Meta-analysisEffect of orally administered N-acetylcysteine on chronic bronchitis — meta-analysisView study →}.

A 2024 systematic review of mucolytics as a class reached the same conclusion — fewer exacerbations and hospitalisations in stable COPD — while noting no improvement in lung function, mortality or quality-of-life scores ^{3Reference 3Ohnishi H et al. · 2024Meta-analysisEfficacy and safety of mucolytics in patients with stable chronic obstructive pulmonary disease — systematic review and meta-analysisView study →}.

The top-tier synthesis agrees but sizes the effect modestly: the Cochrane review of mucolytics versus placebo, pooling 38 trials in over 10,000 participants, found a small reduction in exacerbations and in days of disability, with the caveat that the older, larger-effect trials were at higher risk of bias and that more recent trials show smaller benefit ^{19Reference 19Poole P et al. · 2019Systematic reviewMucolytic agents versus placebo for chronic bronchitis or chronic obstructive pulmonary disease — Cochrane systematic reviewView study →}.

Gap: the benefit is fewer or shorter exacerbations, not better lung function or survival; it applies to chronic mucus-producing disease rather than to healthy lungs; and the effect size has shrunk as trial quality has improved.

A meta-analysis of 22 trials (n≈2,515) in polycystic ovary syndrome found NAC significantly raised progesterone (SMD 0.95), endometrial thickness (SMD 0.58) and — versus metformin — luteinising hormone (SMD 0.67), while estradiol, SHBG and FSH did not change ^{4Reference 4Viña I et al. · 2025Meta-analysisEfficacy of N-acetylcysteine in polycystic ovary syndrome — systematic review and meta-analysisView study →}.

A separate head-to-head meta-analysis found NAC broadly comparable to metformin on several PCOS parameters ^{5Reference 5Song Y et al. · 2020Meta-analysisComparison of the efficacy between NAC and metformin in treating PCOS patients — meta-analysisView study →}.

These are hormonal and ovulation-related surrogate markers; the trials are heterogeneous and often small.

Gap: improvements are in intermediate markers — hormones, ovulation, endometrial thickness — not consistently in live-birth rates, and the underlying trials vary widely in quality.

A 2025 meta-analysis of 5 trials reported that NAC improved semen volume (MD 0.69 mL), sperm concentration (MD 4.43 million/mL), total motility (MD 9.69%) and normal morphology (MD 1.36%) compared with control, with no reported side effects ^{8Reference 8Syarif S et al. · 2025Meta-analysisN-acetylcysteine to improve seminal parameters — systematic review and meta-analysisView study →}. The rationale is antioxidant protection of sperm DNA and membranes against oxidative stress.

Gap: only five small trials, and the outcomes are semen parameters — not pregnancy or live-birth rates, which is what actually matters to couples.

This is where the evidence has genuinely turned. An earlier meta-analysis of 7 RCTs (n=220 on NAC against 220 on placebo) found adjunctive NAC improved PANSS negative and total scores, plus working memory, but only after 24 weeks or more of treatment ^{6Reference 6Yolland CO et al. · 2020Meta-analysisMeta-analysis of randomised controlled trials with N-acetylcysteine in the treatment of schizophrenia — meta-analysisView study →} — suggesting a slow-onset effect. However, a more recent systematic review of 8 trials (594 patients) found no difference from placebo on any PANSS subscale at either time point, and concluded NAC should not be recommended as a routine antipsychotic augmentation ^{7Reference 7Zhang Q et al. · 2024Meta-analysisEfficacy and acceptability of adjunctive N-acetylcysteine for psychotic disorders — systematic review and meta-analysisView study →}.

Gap: the positive signal was time-dependent and has not survived the larger, more recent pooled analysis; the two meta-analyses disagree, so this is an open question rather than an endorsement.

A 2024 meta-analysis of 12 trials (904 patients, NAC 1,000–3,000 mg/day for 8–24 weeks) found a small but significant reduction in depression scores (SMD −0.24, 95% CI −0.44 to −0.05), with the effect concentrated in bipolar rather than unipolar depression ^{9Reference 9Peng TR et al. · 2024Meta-analysisEfficacy of N-acetylcysteine for patients with depression — updated systematic review and meta-analysisView study →}. The confidence interval is wide and close to zero — and a dedicated meta-analysis of NAC in bipolar depression specifically found no effect ^{17Reference 17Pittas S et al. · 2021Meta-analysisThe effect of N-acetylcysteine on bipolar depression — systematic review and meta-analysis of randomized controlled trialsView study →}.

Gap: the effect size is small and the pooled estimate imprecise; benefit appears limited to certain diagnoses, and larger trials are needed to confirm it.

A 2024 meta-analysis of 6 RCTs (n≈195) found adjunctive NAC improved total Y-BOCS scores, but only in a narrow window of 5–8 weeks of treatment — with no significant effect at shorter or longer durations, and no separate improvement in obsession or compulsion subscores ^{10Reference 10Eghdami S et al. · 2024Systematic reviewSafety and efficacy of N-acetylcysteine as augmentation in obsessive-compulsive disorder — systematic reviewView study →}. A broader meta-analysis of glutamatergic augmentation agents including NAC found the class beneficial alongside SSRIs ^{11Reference 11Hadi F et al. · 2021Meta-analysisGlutamatergic medications as adjunctive therapy for moderate to severe obsessive-compulsive disorder — meta-analysisView study →}, consistent with NAC’s action on the cystine–glutamate antiporter.

Gap: the benefit is small, time-window-dependent and rests on very few patients; it has not been replicated in large, long trials.

A 2024 meta-analysis of 11 RCTs found NAC reduced craving ratings across several substances — alcohol, cocaine, nicotine, amphetamine, cannabis — with a moderate pooled effect (SMD −0.61, 95% CI −1.17 to −0.06), but with very high heterogeneity (I² = 85%) and no effect on withdrawal symptoms or abstinence; the authors called the evidence weak ^{12Reference 12Cuocina M et al. · 2024Meta-analysisEffect of N-acetylcysteine on craving in substance use disorders — meta-analysis of randomized controlled trialsView study →}. The mechanistic basis is NAC’s effect on glutamate signalling in reward circuitry.

That last point is the one the highest-tier evidence settles. The Cochrane review of pharmacotherapies for cannabis dependence assessed NAC directly and found no clear benefit for abstinence or for retention in treatment, rating the evidence low-quality ^{20Reference 20Nielsen S et al. · 2019Systematic reviewPharmacotherapies for cannabis dependence — Cochrane systematic reviewView study →} — so the craving signal has now been tested against the outcome that matters and did not carry over.

Gap: craving is a subjective intermediate outcome; the pooled effect is imprecise, heterogeneity is extreme, and the Cochrane assessment found no effect on actual abstinence or relapse.

An umbrella review of meta-analyses (moderate-quality evidence) found NAC associated with less contrast-induced nephropathy (OR 0.72, 95% CI 0.65–0.79) and slightly lower serum creatinine — but with no effect on the outcomes that matter, namely need for dialysis or mortality ^{13Reference 13Zhu R et al. · 2023Meta-analysisAssociation of N-acetylcysteine use with contrast-induced nephropathy — umbrella review of meta-analysesView study →}. Large individual trials have been negative, and the creatinine signal may partly reflect NAC’s known interference with creatinine assays rather than true kidney protection.

Gap: the benefit is confined to a biomarker that NAC can artefactually lower; it does not reduce dialysis or death, and the largest RCTs are null.

In a meta-analysis of hair-pulling treatments, NAC showed a significant benefit versus placebo in adults (SMD −0.75, 95% CI −1.36 to −0.13), but behavioural therapy with habit-reversal training had by far the largest and best-supported effect (SMD −1.22) ^{14Reference 14Farhat LC et al. · 2020Meta-analysisPharmacological and behavioral treatment for trichotillomania — updated systematic review with meta-analysisView study →}. The NAC signal comes from single trials that need replication, and a well-known paediatric trial was negative.

Gap: the effect rests on unreplicated adult trials, does not extend to children, and is smaller than the first-line behavioural treatment.

A meta-analysis of randomised trials concluded NAC is safe and tolerable and reduces hyperactivity and irritability while enhancing social awareness in children with autism spectrum disorder, but cautioned that more evidence is needed before any general recommendation ^{15Reference 15Lee TM et al. · 2021Meta-analysisEffectiveness of N-acetylcysteine in autism spectrum disorders — meta-analysis of randomized controlled trialsView study →}. The trials are few and small, and the effects are on associated behavioural symptoms rather than core features.

Gap: few small trials, targeting irritability and hyperactivity rather than core autism features; the authors themselves stop short of recommending it.

A 2025 meta-analysis in metabolic-dysfunction-associated steatotic liver disease found NAC did not significantly change hepatic steatosis, liver enzymes or lipid profiles — the endpoints that define the disease — but did improve oxidative-stress, glycaemic and inflammatory markers ^{16Reference 16Sinaeinejad M et al. · 2025Meta-analysisEfficacy of N-acetylcysteine on liver function and metabolic profiles in MASLD — meta-analysisView study →}. So the antioxidant mechanism registers in the blood without translating into liver improvement.

Gap: null on every clinically defining liver endpoint; only upstream biochemical markers move, which is not evidence of benefit for the condition itself.