Inside the ProleevaMax Studies: What They Found, and What They Didn't

Two small human studies of ProleevaMax have been carried out. Neither had a control group, neither was randomised, and neither provides evidence that the formula caused any change. The 60-day study did not meet its primary endpoint. The 8-week study reported a positive result on its main inflammation marker using a one-sided test whose prospective justification we cannot establish — read conventionally, that result is around p = 0.08. We are publishing the detail rather than the headline, because the headline would overstate what we have.

Why this page exists

Most supplement brands with a study publish the best number in it.

We have two studies, and if we picked the best numbers out of them we could write something that sounded far stronger than the evidence supports. Several versions of that copy have already appeared on our own site, and we are in the process of removing them.

So this is the alternative: the complete account, including the parts that do not help us. If you are trying to decide whether ProleevaMax is worth your money, the honest input is not our best statistic — it is an accurate picture of how much weight the research can bear.

The short version: these are early, small, uncontrolled studies. They are a reason to keep investigating. They are not a reason to believe the formula has been proven to work.

First, what "uncontrolled" means — because it decides everything else

Both studies gave the product to everyone who took part and measured them before and after. Nobody received a placebo. Nobody was randomly assigned. Everybody, including the researchers, knew what was being taken.

That design cannot separate the formula from everything else happening at the same time:

• Regression to the mean. People join studies when symptoms are bad. Bad patches often improve on their own, and a measurement taken at the worst point tends to look better later regardless of treatment.
• Expectation. Knowing you are taking something intended to help changes how you report feeling. This is not weakness or dishonesty — it is a well-documented, measurable effect, and it is exactly what placebo groups exist to quantify.
• Everything else. Season, activity, sleep, other medication, other supplements, a change in diet.

A before-and-after change in an uncontrolled study tells you what happened. It cannot tell you what caused it. Every number below carries that limit, and no amount of statistical work removes it.

Study one — the 60-day evaluation

Run with Reputable Health as Experiment 486. Clinical Study Report version 1 issued 27 July 2026. Single-arm, uncontrolled, no placebo, no randomisation.

Who was actually in it. This matters more than it sounds, because a single number gets quoted as "the study size" when the real picture is a ladder:

| Stage | People |
|---|---|
| Enrolled | 35 |
| Never started | 2 |
| Full analysis set | 33 |
| Completed the protocol | 27 |
| Evaluable for the inflammation marker | 26 |
| Evaluable for adherence | 30 |

If you see "27" quoted as the study size, that is the completer count, not the enrolment. We have used the wrong rung of this ladder ourselves in the past.

The primary endpoint, which is the one that counts

The study's pre-specified main question was whether whole-group hs-CRP — a general inflammation marker — would fall.

It did not, to any degree that statistics can distinguish from chance. The end-of-study to baseline ratio was 0.872, at p = 0.683.

A p-value of 0.683 means a result at least this large would turn up roughly two times in three by chance alone. The primary endpoint was not met. That is the single most important sentence on this page, and it belongs near the top of any honest summary of this study.

The subgroup figures we are not publishing

Among participants who entered with already-elevated hs-CRP, exploratory analyses showed larger reductions. Two problems:

They are exploratory, not primary. Analyses chosen after seeing the data, in groups of ten and six people, neither reaching statistical significance. Endpoint hierarchy exists precisely so that a study which misses its main question cannot be rescued by a subgroup that looks better.

The report contradicts itself. The synopsis and conclusions section states one pair of percentages; the results table in the body states a different pair, from different baseline values. Those cannot both be right.

So we are not publishing either pair. LanFam is not treating those percentages as settled until Reputable Health reconciles the discrepancy. If you have seen a specific percentage attached to this study on any of our surfaces, it came from this unresolved analysis, and it should not have.

What the participants reported

The questionnaire-based measures did move, and unlike the subgroup work these figures are stable and consistently reported. Among completers, using the PROMIS instruments:

| Measure | Mean change | 95% confidence interval | p |
|---|---|---|---|
| Fatigue | −6.92 points | −9.70 to −4.14 | <0.001 |
| Pain interference | −4.66 points | −7.17 to −2.16 | <0.001 |
| Pain intensity | −0.73 points | — | <0.001 |

Around 70% of completers showed a meaningful improvement in fatigue, and around 53% in pain interference.

Read those with the whole page in mind. They are self-reported, in completers only, in a study with no control group — which is the design most vulnerable to expectation effects, and self-reported outcomes are the measures most sensitive to them. They are consistent with the formula helping. They are equally consistent with what an uncontrolled trial produces when people know what they are taking. We cannot tell you which.

It is also worth noticing the shape of the results: the objective blood marker did not move, and the subjective measures did. That pattern is exactly what you would expect from expectation effects, and exactly what you would expect if the formula affects symptoms more than it affects systemic inflammation. One study of this design cannot distinguish between those explanations.

Study two — the 8-week published study

Published as Proleevamax and its role in management of chronic pain and inflammation: Clinical efficacy and trial outcomes, in the International Journal of Advanced Community Medicine, 2024;7(3):11–14.

Twelve people. Open-label, single-arm, no control, no randomisation, no blinding. Four capsules daily for eight weeks. Mean age 57, range 24 to 76, six men and six women, with mixed chronic conditions.

The journal describes itself as peer-reviewed. We have not independently verified that, and we mention it because "published" and "rigorously reviewed" are not automatically the same thing.

The inflammation marker

Across these twelve uncontrolled participants, average hs-CRP moved from 3.44 to 2.52 — about 27% — reported at p = 0.042. With no control group, that movement cannot be attributed to the formula.

That is the number most often quoted about ProleevaMax, and it needs three qualifications.

It is a one-sided test. A one-sided test asks only whether the marker went down, discarding the possibility of it going up. That is legitimate only when specified in advance and justified. The article states no rationale, and no protocol or statistical analysis plan for this study has been located. Read conventionally, as a two-sided test, the same result is approximately p = 0.08 — above the usual 0.05 threshold. It would also fall short of the one-sided threshold conventionally matched to a two-sided 0.05 test, which is about 0.025.

We are deliberately not saying this study "failed." Without the analysis plan we cannot establish what was pre-specified, and no efficacy conclusion can be drawn from an uncontrolled study of twelve people in either direction. The correct description is that the analysis plan is unresolved.

It is a marker, not a symptom. hs-CRP is a general, nonspecific measure — we explain what it does and doesn't tell you separately.

Four of the twelve got worse. Individual data in the paper show four participants finished with higher hs-CRP than they started with. Group averages hide that, and you should know it.

The pain measure, and the claim that isn't in the paper

Pain was recorded as Present Pain Intensity on a 0 to 3 scale, and the group mean moved from 2.00 to 0.92.

The article reports no p-value for any pain measure. It is descriptive only.

This matters because LanFam copy has previously claimed a "22-point reduction in pain scores" and attached "p = 0.042" to pain. Neither is in the paper. There is no 22-point figure anywhere in it — the scale only runs to 3 — and p = 0.042 is the inflammation-marker result, not a pain result. Both claims are wrong, both came from us, and both are being removed wherever they appear.

Where the paper contradicts itself

Reading it closely, several things do not line up, and you are entitled to know that we noticed:

• The statistics section says a paired t-test; a table note says one-way repeated ANOVA.
• The results text says the marker fell "by 50%"; the table shows 3.44 to 2.52, which is 27%.
• The methods say eight weeks; the pain questionnaire is described as given at start, 30 days and 60 days.
• The text describes the pain scale as running "from 0 to 1"; the table shows 0 to 3.

None of these is fatal on its own. Together they are a fair reason to treat the paper's precision cautiously.

The men-and-women split

Male participants showed a larger reduction than female participants, whose marker was essentially unchanged.

This does not mean the formula does not work for women. Six people cannot support that conclusion. Failing to detect an effect in a group that small is not the same as establishing there is no effect — a distinction that gets lost constantly in supplement marketing, usually in whichever direction suits the seller.

Where that leaves things

Putting both together, without softening:

Neither study provides controlled evidence that the formula works. The 60-day study did not meet its primary endpoint. The 8-week study reported a positive result on its main inflammation marker using a one-sided test whose prospective basis has not been established, and whose two-sided equivalent is approximately 0.08. No causal conclusion is available from either.

What we may honestly say is that two small human studies of the formula exist, that participants in both reported improvement, and that this is a reason to run better research — not evidence that the product works.

What we may not say, and are removing wherever it appears: clinically proven · RCT-backed · placebo-controlled · a significant pain result · a 22-point reduction · any claim that the formula caused a change.

What would actually settle it

A randomised, placebo-controlled, double-blind trial with a pre-specified primary endpoint and a published analysis plan, large enough to detect a realistic effect. Everything on this page describes what happens before that study is run.

Two nearer-term steps: Reputable Health reconciling the conflicting subgroup figures, and the finalised Clinical Study Report from the 60-day evaluation. When those land, this page gets updated rather than quietly replaced.

What you can do with this

If you are deciding whether to try it, the research is not the reason to. The reasonable reasons are the formulation logic, the standardization percentages and the 90-day guarantee that lets you find out for yourself. How the formula is designed sets out the reasoning, and the full ingredient list shows what is in it and at what standardization.

If you take medication, speak to your doctor first — see who should not take boswellia.