Measurement guide

Running A Fair Sixty-Day Test: The Questionnaire, The Placebo And The Threshold

A sixty-day guarantee gives you time to run a proper test on yourself. It does not tell you how. Here is the questionnaire the trials use, how big a change has to be before it counts, and how much of a change you should expect from a placebo.

By VigorHorse Team · Last updated September 26, 2026 · 11 min read · Measurement guide

Why a jar needs a measuring stick

Ask most people at the end of a jar whether it worked and you get an answer built out of mood, memory and the fact that they paid for it. That is the least reliable instrument in the whole category, and the results timeline page says so and offers a simple remedy: write down four numbers on day one and read them again later.

This article goes one level further. It covers three things that page does not: the questionnaire clinical trials use, the size of change that counts as real, and how much improvement people report from a placebo. Together they turn “I think it helped” into something you can hold up against the published record.

One boundary first. The instruments below were built for erectile function, and they have been tested mostly in men who have erectile dysfunction. VigorHorse is a dietary supplement, not a treatment, and its label names circulation, libido and stamina rather than erectile dysfunction. Where a tool does not fit what you care about, this article says so and points you at a plainer one.

The questionnaire researchers actually use

In 1997, Rosen and colleagues published the International Index of Erectile Function, the IIEF, in the journal Urology. It has 15 questions covering five areas: erectile function, orgasmic function, sexual desire, intercourse satisfaction and overall satisfaction. It was translated into ten languages, tested for consistency, and, importantly, shown to be sensitive to treatment: scores changed across all five areas in men who responded to treatment and did not in men who did not.

In 1999, Rosen and colleagues followed it with a short form, the IIEF-5. Five of the fifteen items were kept, the ones best at spotting erectile dysfunction and its severity. In 1,152 men, 1,036 with erectile dysfunction and 116 without, a cut-off score of 21 on a scale of 5 to 25 separated them best, with sensitivity of 0.98 and specificity of 0.88. Scores of 22 to 25 mean no dysfunction and 5 to 7 mean severe.

Trial reports mostly use a third version: the six-item erectile function domain of the full IIEF, which runs from 6 to 30. When a paper says an intervention improved erectile function by so many points, it is usually talking about that domain, or about the short form.

How big does a change have to be?

A trial can call a difference statistically significant when it is too small for anyone to feel. The number that answers the other question, what change matters to the person, is called the minimal clinically important difference.

In 2011, Rosen and colleagues estimated it for the erectile function domain using data from 17 randomised trials of tadalafil, 3,345 patients treated for 12 weeks. They anchored it to a question about how often intercourse was satisfactory. The estimate was 4 points overall, with a sensitivity of 0.74 and a specificity of 0.73, and it varied with how severe the dysfunction was to begin with.

From the 2011 analysis. The authors say the estimate came from drug trials and needs replicating in other drugs and in non-drug interventions.

That caveat is worth taking seriously. It means the threshold is borrowed from a different kind of trial, and the IIEF-5 and the six-item domain are not the same scale. Treat the table as a rough guide to what “small” means, not as a line the law draws.

Hold it against a real supplement trial. A 2009 study in 50 Caucasian men with mild erectile dysfunction gave them 2,400 mg of a maca dry extract or placebo for 12 weeks. IIEF-5 scores rose by 1.6 points on maca and 0.5 on placebo, and the difference between them was statistically significant. The authors call it a small but significant effect. Both things are true: the difference is unlikely to be chance, and it is about a point, which is below the change patients tended to notice in the drug data. The maca article covers the doses behind that result.

The placebo problem, in numbers

Every before-and-after judgement you make on yourself has a built-in comparison that you never see: what would have happened anyway. The placebo arms of large trials give you a sense of it.

A 2020 systematic review in JAMA Network Open pooled 63 double-blind, placebo-controlled trials of prescription erectile drugs, with 12,564 men whose average age was 55. In the placebo arms alone, erectile function improved significantly, with a small-to-moderate effect size (Hedges g 0.35). The effect was larger in men whose dysfunction was linked to post-traumatic stress, at 0.78.

A separate analysis pooled the placebo arms of 42 sildenafil studies, 4,360 men. Of them, 13.5% ended the trial with an erectile function domain score of 26 or higher, a level the authors treat as normalised, while taking nothing active. The strongest predictors included being younger than 45, having mild dysfunction at the start and not having diabetes. The longer dysfunction had lasted, the less likely a placebo response became.

Those trials tested prescription drugs in men with dysfunction, so they do not tell you the exact number for a botanical gummy in a man without a diagnosis. What they do tell you is that “I feel different” can be produced by expectation, by ordinary fluctuation, by the month having been a better one, and by the fact that you started paying attention. A placebo arm bundles every one of those together, and a home test cannot separate them.

What the supplement reviews conclude

Given that background, it is worth reading what the pooled supplement trials found, because they set the honest prior.

A 2024 network meta-analysis compared nutraceuticals for erectile dysfunction across 15 randomised, placebo-controlled trials in 1,000 men. Its verdict was measured: against a background of general ineffectiveness of most nutraceutical interventions, L-arginine and one carnitine combination showed some usefulness, mainly alongside prescription drugs in men with an organic cause. No other regimen reached statistical significance.

A 2026 review of 14 herbal-supplement trials in 1,227 men was more encouraging overall, with a significant pooled effect on erectile function and saffron and ginseng singled out as effective, neither of which is on this jar. Of two names that are, it says the evidence for tribulus terrestris and maca “remains insufficient”, and asks for larger trials with longer follow-up.

None of that says a gummy cannot help anyone. It says that the published trial record is thin for most of what is on a label like this, and that the honest expectation going in is modest. That is the right frame for a test: not “prove it works”, but “find out cheaply whether it does anything for me”.

A fair test you can run on yourself

Nothing here is complicated. It just has to be done in order, and most people skip the first two steps.

1. Name the question before the first gummy. Erections, desire or stamina? Pick one main outcome. Three loosely watched outcomes give you three chances to find a pattern that is not there.
2. Pick the instrument. For erectile function, the IIEF-5 or the domain form from a clinician. For desire or stamina, a dated one-to-ten self-rating, written down rather than remembered. Use the same one throughout.
3. Measure twice before you start, a fortnight apart. This is our suggestion rather than a published rule, and it costs nothing: the gap between your two baselines is your own noise. If they differ by two points, a two-point change later tells you nothing.
4. Hold everything else still. Sleep, alcohol, training and weight all move the same outcomes, and the lifestyle evidence article shows how large some of those effects are. Change one thing, the gummy.
5. Re-measure at day 30 and day 50, writing the new figures underneath the old rather than over them. The results timeline page explains why day 50 is the practical decision point: the guarantee runs 60 days from the date of purchase, and a refund is a phone call and a posted parcel, not a click.
6. Compare against the threshold and against your own noise. If you used the domain form, a mild baseline needs about 2 points, a moderate one about 5, a severe one about 7, and all of them need to exceed the gap between your two baselines.
7. Decide on day 50, not day 60. That leaves ten days of slack for the refund process, which is spelled out on the refund policy page.

A worked example, with invented numbers

The figures below are made up to show the arithmetic. They are not from any trial and not from any customer.

Imagine a man with mild erectile difficulty who scores 17 on the six-item domain in the first week of September and 19 a fortnight later, before any gummy. His baseline wobbles by two points all on its own, which is exactly the size of the mild-dysfunction threshold. He starts the jar and writes down 20 at day 30 and 21 at day 50.

It is tempting to say he gained four points, from 17 to 21. Measured from his second baseline the gain is two, and from the average of the two it is three, while his own fortnight-to-fortnight swing was two. That puts the change at the edge of the mild threshold and only a point beyond his own wobble, which makes it interesting but not decisive. Add the typical placebo response from the drug trials, and the honest sentence is that something may have moved, and that he cannot say what.

Now change one number. Suppose his day-50 score is 19. That is inside his own baseline range, so the test has told him something quite useful at a price of one jar: nothing to see here, and the guarantee is still open. The point of the exercise is that both outcomes are informative, and neither depends on how he feels about having paid for the product.

The same logic works for a self-rating out of ten. If your two baseline readings are 5 and 6, a 6 at day 50 means nothing. An 8 or 9 in the same conditions, with sleep, alcohol and training unchanged, is worth taking seriously, and is still not proof.

What each outcome does and does not tell you

The last column is our reading of the published thresholds and placebo data, not a clinical rule.

The honest ceiling of a home test

You cannot blind yourself to your own gummy. That is the single fact that separates a home test from a trial, and no protocol removes it. What a home test can do is reduce the noise, fix the question in advance and force a decision at a point when the guarantee still works for you.

It is worth being blunt about what that means for the seller's own numbers as well. The distributor's review summary says the commonest wait is four to eight weeks, and that fits the trial timetables, which is why the results timeline page reads it the way it does. It is still a summary figure from buyers, not a measurement, and this website says so on every page that quotes it.

So the goal is modest and worth having. By day 50 you will know whether your own dated numbers moved by more than they wobble, and you will know it while a refund is still a phone call away. If they did not, you have lost a jar and gained an answer. If they did, you have a reason to keep going that is a good deal better than a feeling.

References