Why can I eat cheese but not drink milk?

Lactose content differs a lot between dairy products. Aged hard cheeses contain very little, yogurt and kefir are generally lower than milk because fermentation consumes some of it, and fresh milk is relatively high. So a dose-dependent pattern like this is common and coherent.

How long does it take for lactose to cause symptoms?

For many people discomfort appears somewhere in the first few hours after eating. A symptom that reliably shows up the next day is more likely pointing at something else, which is why logging the time matters.

Should I cut out all dairy to find out?

It is a poor first experiment, because it removes lactose, dairy protein, fat and companion foods all at once and tells you nothing about dose. Varying the form and amount while keeping dairy in your diet usually gives a more useful answer.

Can a food diary confirm lactose intolerance?

No. A diary can surface a possible dose-dependent pattern that is worth investigating, but clinicians have proper tests available, and several conditions produce very similar symptoms. Bring your notes to an appointment rather than settling it yourself.

Am I Lactose Intolerant? A Two-Week Check

Am I lactose intolerant? Before cutting dairy out, run a careful two-week dose observation. How to design it and what it cannot tell you.

Milk in your morning coffee, and by eleven you are bloated. But cheese seems fine. Yogurt seems fine. A milky dessert last weekend was definitely not fine. Somewhere in there is a pattern, and the internet's advice — cut all dairy for a fortnight and see — is the least informative way to find it.

Here is a more careful design: keep dairy in, vary it deliberately, and record enough to see whether your reaction depends on dose. That distinction is what separates a useful two weeks from a confusing one.

"Dairy" hides at least four variables

When people ask whether they are lactose intolerant, they usually mean "does dairy disagree with me". Those are different questions, because a glass of milk brings several things at once.

Lactose. The sugar in milk, which needs an enzyme to be broken down. How much of that enzyme adults produce varies enormously between individuals and populations, and it is normal for it to decline after childhood.

Lactose content varies a lot by product. This is the most useful fact in the whole topic. Fresh milk is relatively high. Yogurt and kefir are generally lower, because fermentation consumes some of the lactose. Aged hard cheeses like mature cheddar or parmesan contain very little. Butter and cream are mostly fat. So "cheese is fine but milk is not" is an extremely common and entirely coherent pattern — and it is a dose story, not a mystery.

Dairy proteins. Casein and whey are separate from lactose. Some people report discomfort that does not track lactose content at all. A true milk allergy is a different thing again and can be serious — that one belongs with a clinician, not a food diary.

Everything dairy travels with. Milk arrives in coffee, so caffeine comes with it. Yogurt arrives with granola, fruit and sometimes a lot of added sugar or sweeteners. Cheese arrives on pizza. Ice cream arrives with fat, sugar and cold. Any of those may be the actual driver.

Why "cut all dairy for two weeks" muddies the answer

The elimination-first approach is popular because it is simple. It is also a poor experiment for this particular question, for four reasons:

You remove four variables simultaneously — lactose, protein, fat and companion foods — and learn nothing about which one mattered.
You lose the dose information entirely. The most actionable answer here is usually "milk is a problem, yogurt is not", and a total removal cannot show that.
Expectation shapes what you notice. Having decided dairy is the problem, a fortnight of feeling better is not clean evidence, and a fortnight of feeling the same often gets explained away.
The reintroduction is unrealistic. Going straight back to a large glass of milk after two weeks off tends to produce a dramatic reaction that says more about the sudden dose than about your usual tolerance.

Elimination is a legitimate tool, and it has its place — usually with professional guidance and a structured reintroduction. It is just the wrong first move here. The mistakes that skew elimination results covers the broader trap.

Design an observation instead

Two weeks. Dairy stays in your diet. You vary the form and the dose deliberately and record what happens.

Four things to log per dairy occasion:

What form, and how much. "200 ml of milk in coffee." "Two thick slices of cheddar." "A small pot of plain yogurt." Rough is fine, consistent is essential.
What came with it. Coffee, cereal, sugar, fruit, bread, alcohol.
What time. You need this to compare against symptoms later.
How you felt at 30 minutes, 2 hours and 4 hours. Lactose-related discomfort typically shows up somewhere in the first few hours, so a symptom that reliably appears the following day is probably pointing somewhere else.

Symptoms worth recording specifically rather than as "bad": bloating, audible gurgling, cramping, wind, loose stools, nausea. Vague notes make the fortnight unreadable — how to keep gut notes specific is a useful companion here.

The dose ladder

Rather than testing dairy as a category, work up a rough ladder. Test one rung most days, keeping the rest of the meal ordinary.

Rung	Relative lactose	What else it brings
Aged hard cheese	Very low	Fat, salt
Butter	Very low	Fat
Plain yogurt or kefir	Lower than milk	Live cultures, protein
Soft cheese, cream	Low to moderate	Fat
Milk in tea or coffee, small amount	Moderate, small dose	Caffeine
A full glass of milk	Higher	Nothing much else
Milkshake, latte, ice cream	Higher, often large dose	Sugar, fat, cold, volume

If discomfort scales up the ladder, that pattern is consistent with a dose-dependent lactose story, and the practical answer is a threshold rather than avoidance. If it appears at the bottom of the ladder as readily as the top, lactose is a less likely explanation and something else in the picture deserves attention.

Confounders that can imitate lactose intolerance

The reason careful design matters: several common things produce almost identical symptoms.

Fat volume. A large amount of cream or cheese slows stomach emptying for many people, which feels a lot like bloating.
Caffeine. Milky coffee is two variables. Try the same coffee black for a few days, and a similar amount of milk without coffee, before blaming the milk.
Other fermentable carbohydrates in the same meal. Onion, garlic, wheat, beans, apples and stone fruit are frequent contributors to the same symptoms. A cheese sandwich with onion is not a clean test.
Sweeteners. Flavoured yogurts and protein shakes often contain sugar alcohols, which are well known for producing wind and loose stools in some people.
Sleep, stress and your cycle. Gut sensitivity shifts with all three for many people. A rough week can turn a food you normally tolerate into a suspect.
Portion and speed. A large meal eaten fast produces discomfort regardless of its dairy content.

You cannot eliminate all of these. You can note them, which is enough to stop yourself drawing a confident conclusion from a confounded week.

Try this for two weeks

Days 1–3: baseline. Eat as usual, log dairy occasions in the four-field format, plus your symptom notes. No changes.
Days 4–6: the low rungs. Hard cheese and butter as your only dairy. Note anything that shows up anyway — that is useful information about your baseline noise.
Days 7–9: yogurt. A plain, unsweetened pot, on its own, away from coffee and away from a big meal. This is the cleanest test in the fortnight.
Days 10–11: small milk dose. Milk in tea, or a small glass, and, separately, the same coffee black on one of the days.
Days 12–13: larger milk dose. A full glass, or the latte you would normally have, eaten with a plain meal rather than with pastries.
Day 14: read it back. Look for a ladder, not a verdict.

A single tracker holding the meal photo, the time and the symptom note keeps this readable, which is what NutriAI is built for — the dairy occasion and the three-hour note sitting on the same timeline instead of in two apps.

Reading your fortnight

Discomfort scales with lactose dose — a possible dose-dependent pattern. Many people land on a workable threshold: milk in coffee fine, a full glass not.
Symptoms on every rung including butter — lactose looks unlikely. Consider fat volume, meal size, or the companion foods.
Yogurt fine, milk bad, cheese fine — the most common shape of all, and consistent with lactose content varying by product.
No pattern anywhere — keep dairy. Two clean weeks of no relationship is a genuinely useful result, and it saves you from removing a whole food group for nothing.
Confounded and unreadable — normal. Run it again in a calmer month rather than forcing a conclusion.

Whatever you find, this is a personal observation, not a clinical answer. It tells you what to bring to an appointment.

When to talk to a professional

This topic has more reasons than most to involve a clinician early, and a few genuine hazards in doing it alone.

Book an appointment rather than continuing to self-observe if you have blood in your stool, unexplained weight loss, ongoing diarrhoea, vomiting, severe pain, signs of anaemia, or a family history of coeliac disease or inflammatory bowel conditions. These symptoms overlap heavily with lactose intolerance and need proper assessment.

Two specific cautions. First, testing for coeliac disease has to happen while you are still eating gluten regularly, so cutting it out beforehand can make the results unclear — do not remove wheat while you are also investigating dairy. Second, any suspected milk allergy, especially in a child, needs professional assessment rather than home testing, because allergic reactions can be severe.

Clinicians also have proper tests available for lactose handling, which your notes cannot replace. Bringing a clear two-week record makes that conversation much faster — how to raise food and symptom questions with your doctor covers how to present it. And if you are considering cutting dairy out long term, that is a good moment to ask about calcium and vitamin D rather than to work it out alone.

The point of the design

A fortnight of careful observation usually gets you to a threshold rather than a ban, and thresholds are much easier to live with than food group exclusions. Same effort as an elimination, considerably more information.

If you would rather run this with the meals, the doses and the symptom notes on one timeline, that is the view NutriAI keeps. How the score works explains what its estimates are before you rely on them, and the best food symptom tracker guide covers what to look for in any tracker you choose.