OBU Health Review

What OBU Health charges

• Sermorelin: $149/mo — the monthly plan figure on the product page, which the seller describes as the same price per vial
• Sermorelin: $379 — the three-month plan on the same product page, described as the lowest rate with the same price at every refill; it is a prepaid total, not a monthly rate
• NAD+: $219 — the figure in the product block and in the site-wide menu, with no period stated beside it
• NAD+: $129/mo — the seller's own comparison table on the same page prints this as its starting price per month for the same product — it contradicts the figure in the product block above, and both are published

Two figures for one product, on one page

The product block for one compound carries a figure with no period attached to it, and the site-wide menu repeats that figure.

Lower down the same page, the seller's own comparison table prints a different, smaller figure for itself under a heading that names a month.

Both describe the same product, on the same page, on the same day, and neither is presented as superseding the other.

Both are recorded here, with their units, rather than reconciled into one number the seller has never published.

A menu that renders on the not-found page

Requesting an address this seller does not serve returns a not-found page that still carries its full four-item priced menu.

That is the test for site furniture, and this menu passes it: it is navigation, not a product record.

It is still the seller's own published price list, so the figures are real — but the product page is the surface to read them from.

Doing that is what surfaced the contradiction, because the menu carries only one of the two figures.

What the published figure actually buys

Four figures are published across its menu and product pages, and two of them describe the same product in different units without either correcting the other.

The documentation question closes quickly: no certificate is claimed, and the attribute sweep found none linked either.

One hundred and thirty-seven of the two hundred and ten sellers here publish a figure somewhere on their own site, and this is one of them.

The host itself was checked in the same pass as everything above: 404 at a nonsense path, same effective path — and the not-found page carries the seller's full priced menu, which is how the menu was shown to be site-wide, which is what any absence recorded here rests on.

What the roster says about these gaps

One hundred and seventy-six of the two hundred and ten sellers here publish no state list at all, and this is one of them, so eligibility is settled inside an intake rather than before one.

Sixteen of the two hundred and ten sellers here name the facility that fills an order; this is one of the one hundred and ninety-four that do not, so there is no facility a reader can look up.

Four of the two hundred and ten sellers here describe lab work as included in what they charge, and this is not one of them.

It declares two of the sixteen compounds this site tracks, and one hundred and one sellers here declare more, so it is one of the narrower menus here.

How we verified this

• Read every page cited on this row as rendered text in a browser-equivalent fetch, with the response decompressed before any character was counted.
• Fired a nonsense-path control at this host in the same run and compared the address requested with the address actually served, so no claim here rests on a status code alone.
• Joined every text node before scanning for a dollar sign, so a figure split across nested elements could not read as an absence.
• Requested an address this seller does not serve and found its priced menu renders on the not-found page, which established the menu as navigation rather than a product record.
• Read both product pages directly and recorded the product-block figure and the comparison-table figure separately, because they disagree.
• Read the comparison table's other columns as the competitors they name rather than as this seller's own prices.
• Recorded the three-month plan as a prepaid total rather than dividing it into a monthly rate.
• Screened this seller in the live federal warning-letter record by entity and by brand, with a known-good positive control and a nonsense negative control fired in the same run.
• Read the company column of every returned record rather than accepting a full-text match, and attached no letter on a name match.
• Swept every attribute on the pages read for a certificate, not only the link targets, because certificates in this market hide in data attributes and lightbox images.
• Re-counted the state-list, pharmacy-naming and lab-inclusion claims against the current roster rather than reusing figures from an earlier pass.