Evidence-Based Natural Remedies: What the Research Actually Shows (Andrographis, Berberine, D-Mannose and More)

A consolidated review of eleven natural compounds, ranked honestly by the quality and quantity of human trial data — from robust to preliminary.

Overview

Natural compounds occupy an awkward space in health optimisation: often dismissed by conventional medicine, often overhyped by supplement marketers. This review takes a third path, surveying the actual trial record for eleven compounds and formulas — distinguishing strong evidence from biological plausibility, and plausibility from wishful thinking. The compounds covered range from andrographis and berberine, backed by hundreds of rigorous trials, to proprietary combination formulas where individual ingredients are studied but the blends themselves are not.


Andrographis: The Strongest Case for Upper Respiratory Infections

Mechanism: Andrographis (Andrographis paniculata) contains andrographolide, a diterpene lactone with demonstrated anti-inflammatory and antiviral activity. It appears to modulate NF-κB signalling and reduce pro-inflammatory cytokine output.

What the evidence shows: This is the highest-confidence compound in the review. More than eight rigorous RCTs totalling over 1,000 participants support its use for upper respiratory tract infections (URTIs). Across those trials, andrographis consistently reduces symptom severity and duration when taken early in the course of illness.

Key caveat: The evidence is for treatment, not prevention. Taking it prophylactically is not what the trials tested. Timing matters: the benefit appears to be front-loaded — start at first symptoms, not after several days.

What to track on MyKine: Illness duration, symptom severity scores logged during URTI episodes. If you're stacking with other immune compounds, note that the combination blends below lack the same trial support.


Berberine: Metabolic Evidence, Genuinely Dense

Mechanism: Berberine is an isoquinoline alkaloid found in several plants (goldenseal, barberry). Its primary metabolic action is AMPK activation — the same pathway targeted by metformin — which increases insulin sensitivity, reduces hepatic glucose output, and modulates lipid metabolism.

What the evidence shows: Over 100 trials across type 2 diabetes, dyslipidaemia, PCOS, and NAFLD. In some trials, berberine's glucose-lowering effect is described as comparable to metformin, though these are not all head-to-head comparisons and effect sizes vary across studies. The breadth of conditions studied is notable and the evidence base is among the largest for any natural compound reviewed here.

Key caveats: "Comparable to metformin in some studies" is not the same as "equivalent to metformin across the board." Not all trials show the same magnitude of effect. Most trials are in people with established metabolic dysfunction, so extrapolation to healthy individuals optimising from baseline is reasonable but not directly tested. Bioavailability is also variable and formulation matters.

What to track: Fasting glucose, HOMA-IR, HbA1c, and lipid panels (LDL, triglycerides). If you're managing pre-diabetes or insulin resistance alongside lifestyle intervention, berberine is the most evidence-dense natural adjunct available. It is not a replacement for lifestyle or pharmaceutical therapy.


NAC (N-Acetyl Cysteine): Broad Utility, Disease-State Caveats

Mechanism: NAC is a precursor to glutathione, the body's primary intracellular antioxidant. It also has direct mucolytic properties (breaking disulfide bonds in mucus), which explains its oldest clinical use in respiratory conditions.

What the evidence shows: More than 100 RCTs supporting use across COPD, depression, NASH (non-alcoholic steatohepatitis), and PCOS. The evidence base is wide rather than deep in any single area — NAC appears to exert benefit across oxidative-stress-driven conditions because glutathione depletion is a common feature of many of them.

Key caveats: Most trials are in specific disease states. Extrapolation to healthy individuals — for example, using NAC to offset the oxidative load of high-dose supplementation or intensive training — is biologically plausible and commonly practised, but is not directly tested in RCTs. The evidence in depression is preliminary relative to the evidence in COPD and liver conditions.

What to track: Liver transaminases (ALT, AST) if you're using NAC for NASH support or high-supplement-load contexts. Symptom-based tracking of respiratory frequency if using for COPD-adjacent purposes.


Alpha-Lipoic Acid: Neuropathy and Metabolic Markers

Mechanism: Alpha-lipoic acid (ALA) is a naturally occurring dithiol compound that functions as a cofactor in mitochondrial energy metabolism and as a direct antioxidant. It regenerates other antioxidants (vitamins C and E, glutathione) and has demonstrated insulin-sensitising properties.

What the evidence shows: Over 80 trials backing use in peripheral neuropathy and metabolic conditions. ALA has European guideline backing for diabetic peripheral neuropathy — one of the stronger regulatory acknowledgements for a natural compound in this review. It also supports glucose control and inflammatory markers in metabolic contexts.

Key caveats: As with NAC, most trials are in disease states (established neuropathy, type 2 diabetes). Head-to-head comparisons with pharmaceuticals are not consistent across the literature. The metabolic and neuropathic evidence bases are somewhat separate — don't assume benefit in one area implies equivalent benefit in the other.

What to track: Peripheral neuropathy symptom scores if relevant. For metabolic use: fasting glucose, HOMA-IR, inflammatory markers (hsCRP). ALA is one of the more relevant compounds if you're female and tracking metabolic biomarkers alongside PCOS management.


Silymarin (Milk Thistle): Liver Enzyme Normalisation, Mixed Broader Picture

Mechanism: Silymarin is a flavonolignan complex extracted from milk thistle (Silybum marianum). It appears to stabilise hepatocyte membranes, inhibit toxin uptake, promote protein synthesis in liver cells, and exert antioxidant and anti-inflammatory effects.

What the evidence shows: 29 trials in liver disease, with overall positive but mixed results. The strongest signal is for liver enzyme normalisation (ALT, AST reduction). Evidence for disease reversal — halting progression of fibrosis or cirrhosis — is less consistent.

Key caveats: Results are not uniformly positive across liver conditions. Silymarin appears most credible as a hepatoprotective adjunct in elevated-transaminase contexts rather than as a treatment for advanced liver disease.

What to track: ALT and AST on regular panels. If you're tracking liver stress from alcohol, high-dose supplements, or NAFLD risk factors, this is a reasonable compound to monitor response with.


Honey: The Cleanest Evidence for Cough

Mechanism: Honey's anti-tussive effects appear to involve demulcent action (coating inflamed mucosa), local antimicrobial properties (hydrogen peroxide production, osmotic effects), and possibly adenosine receptor effects that suppress cough reflex.

What the evidence shows: 14 cough-focused trials with honey outperforming placebo. In the trial landscape for cough remedies, this is a comparatively clean evidence base. Some trials compare honey favourably to common OTC cough suppressants.

Key caveats: Most trials focus on acute, short-duration cough — not chronic cough of specific aetiology. Dosing and honey type vary across trials.

What to track: Minimal tracking burden here. This is a low-cost, low-risk compound with a practical use case. Diabetics should account for glycaemic impact.


D-Mannose: Biological Plausibility, Statistically Incomplete

Mechanism: D-mannose is a simple sugar that, when excreted in urine, may competitively inhibit the adhesion of uropathogenic E. coli to urothelial cells — the primary mechanism by which most UTIs establish. Unlike antibiotics, it doesn't affect the gut microbiome.

What the evidence shows: Six RCTs with biological plausibility, but these trials have not yet demonstrated statistically significant benefits for recurrent UTI prevention. The mechanism is well-understood; the clinical proof is not yet there.

Key caveats: Active UTI should be treated with standard antibiotics — this is not the context for a "natural first" approach. D-mannose is a potential adjunct for prevention, not treatment. The evidence gap is real: plausibility ≠ proven efficacy.

What to track: UTI frequency if you're prone to recurrence. Note whether episodes change in frequency with supplementation, but interpret with the awareness that current RCTs haven't closed the loop statistically.


Ginkgo: Promising but Preliminary for Cognitive Decline

Mechanism: Ginkgo biloba extracts (standardised to flavonoids and terpenoids) are thought to improve cerebral blood flow, inhibit platelet-activating factor, and exert antioxidant effects in neural tissue.

What the evidence shows: Small, short-duration trials showing promise in cognitive decline. The signal is interesting; the evidence base is preliminary.

Key caveats: Larger confirmatory trials are needed before strong conclusions can be drawn. Small-n, short-duration trials in cognitive conditions are a well-known source of positive findings that don't replicate at scale.

What to track: Cognitive function metrics if you're tracking this domain. MyKine users running cognitive performance protocols might log subjective and objective measures, but should treat ginkgo as exploratory.


Proprietary Combination Formulas (Immune Blends and Samsara-Type Mixes)

The review covers at least two proprietary combination formulas — an immune blend and a herbal combination referred to as a Samsara mix. Both include ingredients with individual evidence bases (which may overlap with compounds described above).

Key caveat — and this matters: The individual ingredients are studied. The combinations themselves lack direct RCT data. Efficacy of the blends is inferred from component research, not measured. This is a meaningful evidence gap. Proprietary blends frequently contain ingredients at doses lower than those used in trials, and interaction effects between components are rarely studied.

What to track: If using these, tracking relevant biomarkers tied to the formula's claimed purpose (immune, metabolic, liver) is the only way to generate personal signal. Don't assume the combination is equivalent to the sum of its studied parts.


Practical Takeaway

  • Highest confidence: Andrographis for URTI (take early, not prophylactically); berberine for metabolic markers; NAC across oxidative-stress-related conditions — these have the largest and most rigorous human trial bases.
  • Strong for specific conditions: Alpha-lipoic acid for peripheral neuropathy (European guideline-backed) and metabolic support; honey for acute cough; silymarin for liver enzyme normalisation — evidence is real but narrower.
  • Track what the compound is supposed to move: Berberine → fasting glucose, HOMA-IR, HbA1c, lipids. ALA/NAC → transaminases, glucose, inflammatory markers. Silymarin → ALT/AST. D-mannose → UTI frequency.
  • D-mannose and ginkgo are not there yet: Biological plausibility (D-mannose) and small-trial signals (ginkgo) are promising, not proven. Weight them accordingly.
  • Proprietary blends have an evidence gap by design: Component ingredients may be studied; the formulas themselves are not. Treat them as experimental and track biomarkers to generate your own signal.
  • None of these replace first-line interventions: Lifestyle, standard medications, and clinical assessment come first. These compounds are adjuncts with documented effect sizes in specific contexts — not substitutes.

Sources

  • The Best Evidence-Based Natural Remedies – Part I — Middle Ground Medicine (https://middlegroundmedicine.substack.com/p/the-best-evidence-based-natural-supplements)
  • The Best Evidence-Based Natural Remedies – Part II — Middle Ground Medicine (https://middlegroundmedicine.substack.com/p/the-best-evidence-based-natural-remedies)