SUMMER

SALE

30

%

OFF

your first subscription order. Limited time only.

Milktab
Tier A · Human RCTLactaseDose1993

More enzyme digested more lactose — until the lactose load got too big

A six-thousand-unit dose beat a three-thousand-unit dose in direct proportion to the enzyme fed. Against a much larger lactose load, neither dose did anything measurable.

Lin, DiPalma, Martini, Gross, Harlander & Savaiano. “Comparative effects of exogenous lactase (β-galactosidase) preparations on in vivo lactose digestion.” Digestive Diseases and Sciences 1993;38(11):2022–2027. · PMID 8223076 · DOI 10.1007/bf01297079

Figure 1 · Reported outcomes by dose and lactose load

The same doses that worked against 20 g of lactose did nothing against 50 g.

≈3,000 IU2 capsules
≈6,000 IU4 capsules / 2 caplets
20 g lactose≈1.5 cups of milk
Reduced

Peak and total breath hydrogen “dramatically reduced.” Symptoms significantly less severe.

Reduced further

Total hydrogen significantly below the 3,000 IU arm (P < 0.05) — a gain the paper calls stoichiometric.

50 g lactose≈4 cups of milk
No effect

Modest reduction, not statistically significant. Symptom scores no different from control.

No effect

Same result at double the dose. The load “appeared to overwhelm the ability of either 3000 or 6000 IU.”

The paper reports these outcomes qualitatively and by significance; it does not publish per-timepoint hydrogen curves, so none are shown here.
What the study asked

Do lactase products differ, and does more enzyme do more?

By 1993 several lactase products sat on shelves in different formats — soft gel, caplet, chewable — and nobody had established whether format mattered or whether dosage was the only variable that counted. The authors set out, in their words, “to compare the efficacy of these products and to further establish a dose-response relationship for use.”

They fed Lactogest, Lactaid, DairyEase and a placebo to lactose maldigesters alongside either 20 g or 50 g of lactose, then measured breath hydrogen — the standard marker for lactose reaching the colon undigested — along with symptom scores.

What they found

Dose scaled with effect, in proportion

At the 20 g load, every preparation sharply cut both peak and total breath hydrogen, and symptoms were significantly less severe across the board. The three formats performed equivalently once matched on enzyme units — format was not the variable that mattered. Dose was.

“Four capsules of Lactogest, two caplets of Lactaid, or two tablets of DairyEase (each treatment containing approx 6000 IU) reduced total hydrogen production significantly (P < 0.05) below that observed with two capsules of Lactogest (containing approx 3000 IU) in a stoichiometric manner.”

Lin et al., 1993 — abstract

Then the ceiling. At 50 g of lactose, hydrogen production fell only modestly and not significantly, and symptom scores for bloating, cramping, nausea, pain, diarrhea and flatus were no different from control.

“The 50-g lactose dose appeared to overwhelm the ability of either 3000 or 6000 IU of beta-gal to assist significantly with lactose digestion.”

Lin et al., 1993 — abstract
Figure 2 · Where the tested loads sit against real food

Both of Lin’s conditions sit outside the range a heavy dairy meal actually lands in.

Never tested
Heavy dairy meal ≈ 15–30 g
7 g150 ml milk
12.6 g8 oz milk
20 gTested · worked
50 gTested · failed
Lactose values from IQWiG / InformedHealth (NCBI Bookshelf) and USDA FoodData Central via the National Dairy Council reference-amount table. The 15–30 g band is our estimate for a large milkshake, a substantial bowl of ice cream, or a large milk coffee with dessert.
Study record
Design
Randomized, multicentre, double-blind
Doses tested
≈3,000 IU · ≈6,000 IU
Lactose loads
20 g · 50 g
Primary endpoint
Breath hydrogen (peak + total)
Secondary
Symptom severity scores
Products
Lactogest · Lactaid · DairyEase
Enzyme units
IU — not FCC
Journal
Dig Dis Sci 38(11):2022–27
Participants
Not stated in abstract
Funding
Not stated in abstract
Our reading

What we take from this study

What it supportsHolds up

This is the clearest published evidence that lactase dose is not a threshold you either clear or don’t. Within the range tested, effect scaled with the quantity of enzyme — stoichiometrically, which is a strong word for a paper to choose. And the study went looking for that relationship rather than stumbling onto it.

It also settles a question people still ask: chewable, caplet and soft gel performed the same once matched on units. Format is not what separates lactase products. The amount of enzyme is.

What it doesn’t showLimits

  • The abstract does not state how many people were studied. We flag that rather than quote a number we cannot source.
  • Doses are in IU, not FCC units. Different measurement systems — we don’t convert between them.
  • The dose-response step is 3,000 to 6,000 IU. Nothing above 6,000 IU was tested, so the curve past that point is unmapped.
  • The primary endpoint is breath hydrogen, not how people felt. Symptom scores moved too, but they are secondary.

Why this shaped milktabOur formula

This study is the reason we didn’t stop at 9,000 FCC. It shows the direction of the evidence plainly: within the range anyone has tested, more enzyme digested more lactose, and a big enough serving of dairy defeated the doses that were available in 1993.

milktab contains 27,000 FCC units of lactase. Lin’s strongest arm was roughly 6,000 IU. Those are different unit systems and we won’t pretend otherwise by converting one into the other — but the shape of Figure 2 is the point. Standard doses were shown to work against 20 g and to fail against 50 g, and the dairy people actually reach for a pill to handle sits in between, where nobody has looked. We built for that gap.

Written by
milktab research team
Reviewed by
Pending
Last reviewed
Record verified
Europe PMC · abstract
Related studies

Others on dose and formulation

The dose behind the research

27,000 FCC units of lactase in one tab

Three times what’s in Lactaid Fast Act, and more lactase than any dose used in the trials on this page.