Evidence check

The Placebo Problem: Why A Good Week Two Proves Little

Somebody starts a supplement, feels better within a fortnight, and concludes the supplement worked. Sometimes it did. But in the trials that test erectile function, the men taking the dummy pill improve too, often by a margin nobody would have predicted. Understanding how large that placebo response is, and what else hides inside a good fortnight, is the most useful thing a buyer of any capsule can learn before the first bottle arrives.

By HoneyPower Team · Last updated September 26, 2026 · 11 min read · Evidence check

What a placebo response actually is

The word placebo gets used as if it meant nothing happened. It means the opposite. A placebo is an inert substance given so that a real one has something honest to be compared against, and it has turned out to be an intervention in its own right. A 2022 review in Nature Reviews Urology on the placebo and nocebo effects in functional urology puts the mechanism simply: the placebo effect is partly the result of the recipient's positive expectations about their health. The nocebo effect is the mirror image, where negative expectations from a substance lead to poor outcomes or adverse events.

That review notes that randomised trials in functional urology have demonstrated the importance of both across several conditions, among them overactive bladder, urinary incontinence and interstitial cystitis, and male and female sexual dysfunction. The authors' argument is that understanding the placebo-nocebo complex could help clinicians maximise the good effects while minimising the bad ones. For the purposes of this article the point is simpler. In the area this product is sold into, the expectation of improvement is itself a measurable force.

It is worth saying what this is not. A placebo response is not fakery, and it is not an accusation. A man who feels better because he expects to feel better does feel better. The problem is only with the inference that follows: that the capsule caused it. That inference needs a comparison, and in a home experiment nobody is running the comparison.

How big it is in erectile function trials

Until recently, the size of the effect in this specific area was surprisingly poorly described. Stridh and colleagues set out to quantify it, and their approach was to go where the placebo arms are. They searched four databases for double-blind, placebo-controlled randomised trials of PDE5 inhibitors, the prescription drug class, published from 1998 to 2018, and pooled the placebo arms.

They analysed 63 studies that included 12,564 men, with a mean age of 55 (range 36 to 68). The main outcome was improvement in the erectile function domain of the International Index of Erectile Function questionnaire in the placebo arm. Erectile function was significantly improved among men taking placebo, with a small to moderate effect size (Hedges g 0.35, standard error 0.03, P below .001). The effect was larger, at 0.78, in men whose erectile dysfunction was associated with post-traumatic stress disorder.

Two things are worth pausing on. The first is that this is a pool of 63 trials of a real prescription drug class, and on a pooled basis the placebo arms improved. The second is the direction of the finding about post-traumatic stress: the men with the most psychological load in the picture responded most to a dummy pill, which fits with expectation and mood doing part of the work, although the abstract does not test that.

Put those together and you get a baseline expectation for any home experiment. Some improvement should be expected from the act of starting something, before the ingredients enter the discussion.

A supplement trial where both arms improved

The clearest illustration in the supplement literature is small, and its authors reported it with admirable plainness. Santos and colleagues took 100 men who had presented with erectile dysfunction, selected 30 who met strict criteria, and randomised them in a double-blind design to tribulus terrestris or placebo. The men were aged 40 or over and non-smokers, with no prostate cancer treatment, no dyslipidemia, no phosphodiesterase inhibitor use and no hormonal manipulation. Any hypertension or diabetes had to be controlled.

The tribulus group received 800 mg a day, in two doses, for 30 days, and the control group received placebo in the same way. Both the erectile function questionnaire and total testosterone were measured before randomisation and after the 30 days.

Mean scores from the trial's abstract. Time changed both groups significantly (P = .0004); the difference between the groups was not significant (P = .7914).

Read the two rows together. Both groups improved by about two points in a month, and the statistical test confirmed that time had an effect. The test between groups found none. Total testosterone did not differ either: it was 409.3 ng/dl after tribulus and 466.3 ng/dl after placebo. The authors concluded that at the dose and interval studied, tribulus was not more effective than placebo on erectile symptoms or on serum total testosterone.

Now imagine that the placebo arm had not existed. A man in the tribulus group would have seen a score rise from 13.2 to 15.3, felt a bit better, and had a perfectly good reason to attribute it to the capsule. A trial of that shape would have looked like a success. The second row is what stops the mistake, and it is exactly the row that nobody has when they buy a bottle and try it.

A fair caveat: 30 men is small, and a trial that small can miss a modest true effect. The lesson here is about the size of the placebo response, not a final verdict on tribulus. The tribulus and testosterone article covers the wider trial record.

A supplement trial where they separated

To be fair to the ingredients, some trials show the opposite pattern, and it is worth seeing what that looks like. Hong and colleagues ran a double-blind, placebo-controlled crossover study of Korean red ginseng in 45 men with clinically diagnosed erectile dysfunction. Each man had 8 weeks on treatment, a 2-week washout and 8 weeks on the other treatment. The ginseng dose was 900 mg three times daily.

The mean questionnaire scores told a different story from the tribulus trial. As the abstract reports them, baseline was 28.0, the ginseng period reached 38.1 and the placebo period 30.9, and the difference between ginseng and placebo was significant (P below 0.01). So the placebo period rose by about three points, which is a genuine placebo response, and the ginseng period rose by about ten. In the global efficacy question, 60 percent of the men answered that ginseng had improved their erections. Penile tip rigidity measured on a RigiScan device also improved for ginseng against placebo.

Two points follow. That trial is what a real separation looks like: the placebo arm moved and the active arm moved further. And even here, the placebo response was not zero. A man who took the placebo and no ginseng would still have reported a better score. Note that this was one 45-man trial at a very large daily dose, 2,700 mg in total. The wider picture, including how the pooled evidence for ginseng looks, is set out in the article on what a ginseng dose actually looks like.

The other things hiding in a good fortnight

The placebo response is one item on a longer list of reasons a person can feel better after starting a supplement without the supplement being the reason. None of the following needs a citation, because they are properties of how people behave and how numbers move. They still deserve a sentence each.

• Starting when things are worst. People tend to buy a product after a bad stretch. Any measure that fluctuates is likely to look better a few weeks after its low point, whatever was taken in between.
• Changing several things at once. A man who decides to sort out his health usually also walks more, drinks less and sleeps earlier. Those changes are real, and any of them may be the true cause of an improvement that gets credited to the capsule.
• Attention. Watching a symptom closely changes how it is reported. So does reducing the anxiety that comes from having done nothing.
• Time. Some things resolve on their own, and a supplement started during a resolving spell inherits the credit.

All of these push in the same direction as the placebo response, which is why a good week two is such a weak piece of evidence. It is not that the capsule cannot be working. It is that the week-two picture would look the same if it were not.

What controlled trials of supplements found overall

Because placebo-controlled trials are the only kind that separate these effects, it helps to see what the pooled controlled evidence says about supplements as a group. Barbonetti and colleagues ran a network meta-analysis of nutraceutical interventions for erectile dysfunction, searching four databases for randomised placebo-controlled trials. Fifteen trials with 1,000 men met their strict criteria, and the outcome was the standardised difference in the questionnaire score.

Against placebo, only a few regimens showed a statistically significant improvement: the combination of propionyl L-carnitine, acetyl L-carnitine and sildenafil ranked highest, followed by L-arginine with tadalafil, sildenafil alone, tadalafil alone and L-arginine on its own. No other treatment regimen showed efficacy with statistical significance. The authors' summary was that, against a background of general ineffectiveness of most nutraceutical interventions, L-arginine and the carnitine mix appeared of some usefulness, especially combined with a prescription drug in organic erectile dysfunction.

None of the six HoneyPower ingredients is L-arginine or a carnitine, and this network analysis is not a verdict on the six. What it does is show the shape of the field when the placebo is taken seriously: a small number of signals, most in combination with a real drug, and a large number of nulls. That is a useful thing to know before deciding how much a good fortnight should count.

How to run a fair test on yourself

You cannot blind yourself, and you cannot run a placebo arm at home. You can make the experiment better than most.

1. Write a baseline first. Before the first capsule, record where things stand on whatever you care about, in a form you can repeat. A widely used questionnaire exists for this purpose; the placebo meta-analysis above reports on its erectile function domain. Ask a doctor which version fits you.
2. Change one thing. If you also plan to start walking or cut back on alcohol, either do those first and take a baseline after they settle, or accept that you will not be able to credit the capsule.
3. Pick the span in advance. Decide before you start when you will look again. The seller publishes a typical wait of 3 to 6 weeks across 1,140 written reviews, which is the distributor's own summary rather than a measurement. The ginseng and maca trials this website cites ran 8 to 12 weeks. A single bottle is about a month, so three bottles matches the research.
4. Look at the same time, the same way. Same scale, same interval. Do not adjust the yardstick when the result is disappointing.
5. Use the guarantee window for what it is for. The 180 days from purchase exist so that a fair look at three months is possible. The guarantee page explains how the window is counted and how to claim it.

None of this guarantees a clear answer. It converts a vague feeling into a comparison, and a comparison is the raw material of every trial in this article.

Five questions to ask of your own good week

• What was my score or pattern before I started, and did I write it down?
• What else changed in the same fortnight?
• How long did the research on this ingredient run before anybody claimed a result?
• If I had been taking a sugar pill, would I expect to feel roughly this way?
• Is the honest answer that I do not yet know?

The last question is the most useful and the least comfortable. A good week two is a reason to carry on, not a reason to conclude. The results timeline lays out what to record and when, and the article on reading a label that prints no amounts explains why this particular bottle needs more of that discipline than most.

References